JP5672596B2 - Conformational homology evaluation apparatus and evaluation method, and structure pattern analysis apparatus and analysis method - Google Patents

Description

  The present invention relates to an apparatus and method for simply and more precisely evaluating structural changes of molecules related to drug efficacy, safety, and the like, by comparing their conformational homology.

  With the progress of post-genomic research, biopharmaceuticals including antibody drugs are being developed in the pharmaceutical industry. This antibody drug has a very deep relationship with proteins, and since artificially designed substances with structures different from those that exist in nature, the establishment of a quality control method that guarantees further safety has been established. It has been demanded. For example, although there is a freeze-drying step in the liquid preparation process of an antibody drug, there is a demand for a means for easily confirming that the structure of the antibody drug has not changed in the solution before and after the freeze-drying (Non-patent Document). 1).

  On the other hand, from the viewpoint of drug discovery, it is important to understand the interaction between proteins and ligands, so many examples of X-ray crystal structure analysis of proteins have been reported, some of which include Japan Protein Structure Data Bank (PDBj) and others. The mechanism of protein structural changes due to interaction with ligands has begun to be elucidated in conjunction with recent advances in graphical interfaces (http://www.pdbj.org/).

However, in X-ray crystal structure analysis, only one of the stable structures is observed. Actually, the function is expressed in the liquid phase, and the conformational change based on the equilibrium occurs constantly.
In other words, there are fluctuations in the three-dimensional structure of the protein in the liquid phase, but these fluctuations could not be expressed simply using an amino acid sequence, an orthogonal coordinate system, or a Z-matrix.

  In particular, from the viewpoint of quality control that guarantees safety, a tool for clearly distinguishing conformational changes from conformations and irreversible structural changes was required.

  The present inventors have found that infrared circular dichroism (VCD) bands of substances having chirality such as biomolecules are extremely sensitive to conformation (Non-patent Documents 2-5). , A new conformational analysis method (Patent Document 1), and a coding apparatus and method for easily expressing the conformation of an organic molecule, based on the idea of the simplified model molecule Proposals have been made (Patent Documents 2-4).

  Among them, in order to solve the problem of the fluctuation of the molecular structure in the liquid phase, 360 degrees are divided into 6 parts and attached with symbols, and each part is further divided into 2 parts and attached with other parts. Although structural homology was evaluated, it was found that there was a problem that a simple arithmetic processing based on coincidence of codes could not completely cope with the problem of fluctuation of molecular structure.

  In addition, when comparing the structural homology between the optimized structural molecular model obtained from the density functional calculation and the molecular model obtained from the X-ray crystal structure analysis, the position of the hydrogen atom is determined in the X-ray crystal structure data. In the dihedral angle classification code involving hydrogen atoms whose coordinate positions are not determined by the above coding method, structural information that can be compared with structural homology is included, but the structural information is sufficient. It has been found that there is a problem that cannot be utilized (Non-Patent Document 1).

  Furthermore, there are cases in which the structural homology is very high even though the amino acid sequence is different when the compound to be analyzed is a protein or peptide, but depending on the conventional conformational analysis method, the amino acid sequence pattern may be detailed. Cannot be extracted in conjunction with a conformational notation. Therefore, it has been desired to construct an effective extraction method for amino acid sequence patterns having a characteristic molecular structure while minimizing the influence of fluctuations in the molecular structure. In particular, an analysis method for determining and extracting a fragment region showing a characteristic molecular structure pattern such as α-helix type or β-sheet type has been desired. Similarly, single-stranded RNAs are also known to have various characteristic three-dimensional structures, and since there are abundant X-ray crystal structure data, even single-stranded polynucleotides such as these RNAs can be used. It is expected that the same analysis method can be applied.

  Thus, in order to solve these problems, an evaluation technique for computer processing of conformational homology has been demanded.

Japanese Patent No. 3991106 WO2008 / 117570 (PCT / JP2008 / 051673) JP 2009-265855 A US-2009-0248321-A1 (US / 12/385856)

Hiroshi Izumi, Farm Tech Japan, 2009, 25, 19 H.Izumi et al., J.Am.Chem.Soc., 2004,126,194 H.Izumi et al., J.Org.Chem., 2007,72,277 H.Izumi et al., J. Org. Chem., 2008, 73, 2367 H.Izumi et al., J.Org.Chem., 2009,74,1231

  The present invention has been made in view of the actual situation and demand of the prior art, and even when simple arithmetic processing based on coincidence of codes cannot completely cope with the problem of fluctuation of the molecular structure, the three-dimensional configuration can be easily performed. An evaluation method capable of comparing the homology of the locus structure, and comparing the molecular model obtained from the X-ray crystal structure analysis without the position information of the hydrogen atom and the molecular model obtained from the theoretical calculation Evaluation method, conformational homology evaluation that can extract amino acid sequence patterns with very high structural homology together with detailed conformational notation even though the amino acid sequence is different when the analysis target compound is protein or peptide Method and apparatus are provided, and the code structure is used to evaluate fragment structure patterns such as α-helix type and β-sheet type. It is an object of the present invention to provide a structural pattern analyzing method and apparatus capable of analyzing it.

In order to solve the above-mentioned problem, a new technique for applying X-ray crystal structure analysis data to structural homology comparison by a conformation homology evaluation apparatus and an evaluation method (Patent Document 4, etc.) previously developed by the present inventors. Priority rules have been created so that even when there is no position information of hydrogen atoms such as X-ray crystal structure analysis data, it can be uniquely described.
This makes it possible to compare the three-dimensional structure data on a huge amount of chemical substances including various peptides and proteins.
Then, a set having a dihedral angle classification code of a bond position corresponding to each conformational structure of each molecule constituting the chemical substance as an element is created, and the elements of the set are consecutive in the dihedral angle classification code. Are included in three adjacent dihedral angle classification codes, and at least one element of the set is a dihedral angle classification code in the center of the three adjacent dihedral angle classification codes. When the above condition is satisfied, the “swing tolerance calculation analysis method” is developed in which it is determined that the homology is high and the homology is low otherwise, and the comparison is performed according to the analysis method.
As a result, even if the three-dimensional structure is the same, it is possible to evaluate almost exactly what the conventional evaluation method (Patent Document 4, etc.) could not perform an appropriate homology evaluation due to “sway” in the liquid phase. became. In addition, it can be proved that there is a position where the conformation changes greatly in the framework region other than the CDR region in the antibody (immunoglobulin) and in the constant region, and it is essential to determine the conformation change of the normal prion and the abnormal prion. The area could be indicated by text information.
Furthermore, a part of the “sway tolerance calculation analysis method” has been improved, and a fragment corresponding to a region of a characteristic molecular structure of a biopolymer composed of many fragments such as amino acid fragments and nucleotide fragments, for example, a protein Dihedral angles with the highest frequency of appearance in each of the dihedral position codes constituting the encoded conformational notation of only the α helix fragment, β helix fragment, and stem fragment characteristic of RNA double helix structure Take out the classification code, use it as a template to refer to the conformational code corresponding to these fragments, and apply the improved “swing tolerance calculation analysis method” to the region to be analyzed for structural homology comparison The fragment with the respective structure pattern It was possible to separate and extract from the fragment.
The present invention was completed by obtaining the above knowledge.

Specifically, the present invention firstly includes:
The processing unit accepts input of a molecular model of the conformation structure of each of a plurality of analysis target compounds,
Based on the molecular model for which the processing unit has received the input, the processing unit includes a three-dimensional configuration composed of a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. Note that the conformation notation is necessary for a structure in which the conformation can be uniquely determined by one conformation notation by means of the conformation notation (hereinafter referred to as an encoded conformation notation). Extracting the encoded conformation notation that leaves the notation of the chemical bonding site and omits unnecessary notation from the molecular model, and stores the extracted encoded conformation notation in the storage unit,
The processing unit retrieves a plurality of the encoded conformational representations from the storage unit;
At each position in the corresponding dihedral angle position, the dihedral angle classification code on both sides is extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrast elements of the set at each position in each analyte compound;
The condition that at least one element among the elements of the set belonging to each compound is common to all compounds, and that the common element is the central element in the set belonging to at least one compound. Search for whether or not
Provided is a conformation homology evaluation method characterized by repeating an operation for performing a calculation process for determining that the homology is high when it is satisfied and the homology is low otherwise.

Second, in the first method,
According to the standard rule for determining the dihedral angle classification code based on the IUPAC nomenclature, the coordinate position is at least one conformational structure of the compound to be compared is a molecular model by X-ray crystal structure analysis. A conformation characterized by selecting the coded conformational notation according to the following priority rules (A) to (C) when a dihedral angle classification code involving a hydrogen atom cannot be determined because it is not defined: A method for assessing locus homology is provided.
(A) In accordance with the rules of IUPAC nomenclature regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) In the IUPAC nomenclature related to conformation, (2) When only one is different from the other, [2] When hydrogen is selected according to the rule of selecting itself regardless of the order The hydrogen atom that should be present is placed in front of the paper, the bond of the dihedral angle position code is placed perpendicular to the paper surface on the other side of the paper, the surface formed by the bond of the hydrogen atom and the dihedral angle position code and the substituent or Select the substituent or atom whose dihedral angle due to the plane formed by the bond of the atom and dihedral angle position code is the smallest clockwise when viewed from the hydrogen atom,
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.

Thirdly, in the first or second method,
When the target compound to be compared is a protein or peptide, the encoded conformation corresponding to the amino acid sequence is selected based on the main chain that binds each amino acid residue and using each amino acid residue as a prefix. Provided is a conformation homology evaluation method characterized by performing structural homology evaluation only on the main chain.

Fourth, in the third method,
A conformation homology evaluation method characterized by extracting the encoded conformation notation and amino acid sequence pattern having high structural homology but different amino acid sequences between proteins or peptides to be analyzed. provide.

In addition, fifthly, in any one of the first to fourth methods,
In the case where the analysis target compounds to be compared have the same fragment, the notation of the dihedral angle position code including the prefix of the fragment is completely the same, and one of them can be selected as a reference compound In this case, the dihedral angle classification code of the conformation structure of the compound is the central code among the three consecutively adjacent ones, and the dihedral angle at the corresponding dihedral angle position of the compound for which structural homology evaluation is performed If all the elements of the set of angle classification codes match the median value of the standard dihedral angle classification code or the dihedral angle classification code adjacent to it, the homology is high. Provided is a method for evaluating conformation homology, which comprises performing determination and performing arithmetic processing.

Further, according to the sixth aspect of the present invention,
A conformation homology evaluation apparatus comprising a display unit, an input unit, a processing unit, a main storage unit, an output unit, an I / O unit, a database unit, and a bus connecting each unit,
In the input unit, in accordance with the input operation, the molecular model of the conformation structure of each of the plurality of analysis target compounds is extracted from the database unit and sent to the processing unit,
In the processing unit, for each molecular model that has received an input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, each of the encoded conformational representations corresponding to the plurality of analysis target compounds is extracted from the storage unit,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrast elements of the set at each position in each analyte compound;
The condition that at least one element among the elements of the set belonging to each compound is common to all compounds, and that the common element is the central element in the set belonging to at least one compound. Search for whether or not
If it satisfies, it is determined that the homology is high, otherwise it is determined that the homology is low,
Subsequently, the same operation at the next position is performed, and when the operation for all corresponding positions is repeated and sent to the output unit,
The output unit provides a conformation homology evaluation apparatus that outputs a processing result generated by the processing unit .

Seventhly, in the sixth device,
In determining the encoded conformational notation for each molecular model in the processing unit, at least one of the molecular models that has received the input is a molecular model by X-ray crystal structure analysis, and IUPAC nomenclature When the dihedral angle classification code involving hydrogen atoms cannot be determined because the coordinate position is not determined according to the standard rule for determining the dihedral angle classification code in accordance with the above, the following priority rules (A) to (C) are followed. There is provided a conformation homology evaluation apparatus characterized by providing a step of selecting the encoded conformation notation.
(A) In accordance with the rules of IUPAC nomenclature regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the surface formed by the bond is the smallest clockwise when viewed from the hydrogen atom,
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.

Eighthly, in the sixth or seventh device,
In the case where the analysis target compounds to be compared have the same fragment, the notation of the dihedral angle position code including the prefix of the fragment is completely the same, and one of them can be selected as a reference compound In this case, the dihedral angle classification code of the conformation structure of the compound is the central code among the three consecutively adjacent ones, and the dihedral angle at the corresponding dihedral angle position of the compound for which structural homology evaluation is performed If all the elements of the set of angle classification codes match the median value of the standard dihedral angle classification code or the dihedral angle classification code adjacent to it, the homology is high. Provided is a conformation homology evaluation apparatus characterized by performing an arithmetic processing for determination.

Ninth,
The processing unit accepts an input of a molecular model of a conformation structure of a fragment corresponding to one or more reference structural patterns,
The processing unit describes the dihedral angle position code corresponding to the chemical bonding site of interest for the molecular model that has received the input, and 360 degrees is divided into 12 for each dihedral angle position code. By applying the encoded conformation notation that combines the specified dihedral angle classification codes, the conformation regarding each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
The extracted encoded conformation notation is stored in advance in the storage unit as a reference structure pattern,
Separately, accepting the input of the molecular model of the conformational structure of the compound to be analyzed,
In the processing unit, for the molecular model of the analysis target compound that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Next, an extracted conformational notation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for the structural pattern analysis is extracted,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, the encoded conformational notation of the analysis target compound is taken out from the storage unit, and the reference structure pattern to be compared is selected from the reference structure patterns stored in the storage unit in advance. , Taking out the encoded conformational notation that matches the structure pattern as a reference for comparison,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrasting elements of the set at the positions in the structural pattern to be analyzed and the reference structural pattern,
Search for whether or not the condition that at least one element is common to the reference structure pattern among the elements of the set to which each dihedral angle classification code belongs,
If it satisfies, it is determined that it matches the reference structure pattern, and if it does not, it is determined that it corresponds to the other structure pattern.
Subsequently, a structure pattern analysis method is provided, in which an operation for performing the same arithmetic processing at the next position is performed, and the operations for all corresponding positions are repeated.

Tenth,
A conformation homology evaluation apparatus comprising a display unit, an input unit, a processing unit, a main storage unit, an output unit, an I / O unit, a database unit, and a bus connecting each unit,
In the input unit, in accordance with an input operation, accepts an input of a molecular model of a conformation structure of a fragment corresponding to one or more reference structure patterns,
The molecular model of the conformational structure of each structural pattern is taken out from the database part and sent to the processing part,
In the processing unit, for each molecular model of each structural pattern that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
The extracted encoded conformation notation is stored in advance in the storage unit as a reference structure pattern,
Separately, accepting the input of the molecular model of the conformational structure of the compound to be analyzed,
In the processing unit, for the molecular model of the analysis target compound that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, the encoded conformational notation of the analysis target compound is taken out from the storage unit, and the reference structure pattern to be compared is selected from the reference structure patterns stored in the storage unit in advance. , Taking out the encoded conformational notation that matches the structure pattern as a reference for comparison,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrasting elements of the set at the positions in the structural pattern to be analyzed and the reference structural pattern,
Search for whether or not the condition that at least one element is common to the reference structure pattern among the elements of the set to which each dihedral angle classification code belongs,
If it satisfies, it is determined that it matches the reference structure pattern, and if it does not, it is determined that it corresponds to the other structure pattern.
Subsequently, an operation for performing the same calculation process at the next position is performed, and when the operation for all corresponding positions is completed repeatedly, it is sent to the output unit,
The output unit provides a conformation homology evaluation apparatus that outputs a processing result generated by the processing unit .

According to the present invention, it is possible to easily perform conformity comparison and structural pattern analysis of conformational structures even when simple arithmetic processing based on code matching cannot completely cope with the problem of fluctuation of molecular structure. it can.
That is, according to the present invention, with respect to one dihedral angle classification code at a corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code has high homology, and the remaining two dihedral angle classification codes are the same. Since the face angle classification code is judged to have low homology, it is processed, so it is possible to easily compare the homology of conformational structures that was impossible with simple arithmetic processing by matching the codes. it can.

  Further, according to the present invention, the molecular model obtained from the X-ray crystal structure analysis without the position information of the hydrogen atom can be compared with the molecular model obtained from the theoretical calculation, and the analysis target compound is a protein. Alternatively, an amino acid sequence pattern having a very high structural homology can be extracted together with a detailed conformation notation in the case of a peptide, although the amino acid sequence is different. Furthermore, fragment structure patterns such as α-helix type and β-sheet type can be extracted with a simple operation using a code. Similarly, in various RNAs such as ribosomal RNA and siRNA, the structure pattern of fragments forming secondary structures such as stem structures and loop structures can be extracted. In addition, by applying this method, a secondary structure that is commonly found between various proteins with low homology at the amino acid sequence level or between various RNAs with low homology at the base sequence level is formed. It is also useful for finding such new structure patterns.

  Therefore, in the present invention, for example, there is a lyophilization step in the liquid preparation process of an antibody drug, but the structural conformation of the antibody drug does not occur in the solution before and after the lyophilization, the encoded conformation notation It can be easily confirmed by structural homology comparison using. In addition, since the present invention can be used for tracking structural changes of proteins involved in function expression, it can be used for designing new drugs by analyzing structural homology and function of orphan receptor proteins.

It is a system block diagram of the conformation homology evaluation apparatus of this invention. The meanings of the numbers in the figure are as follows. 1 display unit, 2 input unit, 3 processing unit, 4 main storage unit, 5 output unit, 6 external interface (I / O) unit, 7 database, 8 bus It is a flowchart of the analysis process which concerns on this invention. It is a figure which shows the definition of the dihedral angle position code corresponding to the amino acid sequence of a protein molecule. It is a figure which shows the classification | category of the code | symbol showing the dihedral angle classification code used for an encoding conformation description. Conformation is uniquely described using symbols such as σ (cis) and τ (trans) based on the relative positional relationship when multiple molecular models correspond in one encoded conformation. It is a figure which shows the illustration which was made possible. It is a figure used as an example for demonstrating the priority rule in case the some encoding conformation description is attained in one molecule | numerator model. It is a figure which shows the molecular model of X-ray crystal structure data 2imm. FIG. 3 is a view showing the difference in structure between 2imm and 2mcp in amino acid sequences 14-15. It is a figure which shows the molecular model of X-ray crystal structure data 1a7o. It is a figure which shows the difference in structure of 1a7o and 3iy2 in amino acid sequence 76-77. It is a figure which shows the molecular model of X-ray crystal structure data 1kid. It is a figure which shows the molecular model of X-ray crystal structure data 1srv. It is a figure which shows the difference in structure of 1srv and 1kid in the amino acid sequence 224-225. It is a figure which shows the molecular model of X-ray crystal structure data 6gst_I. It is a figure which shows the molecular model of glutathione (I) calculated | required from theoretical calculation. It is a figure which shows the molecular model of X-ray crystal structure data 1i4m. It is a figure which shows the molecular model of the prion protein part of X-ray crystal structure data 2w9e. It is a figure which shows the difference in the structure of 1i4m and 2w9e in the amino acid sequence 141-143. It is a figure which shows the difference in structure of 1i4m and 2w9e in amino acid sequence 190-198. It is a figure which shows the molecular model of X-ray crystal structure data 2w9e in which prion protein and immunoglobulin interact. It is a figure which shows the difference in structure of 2mcp and 2w9e in amino acid sequence 150-158. It is a figure which shows the molecular model of the X-ray crystal structure data 2w9e of the area | region where the extracted prion protein and the immunoglobulin light chain interact. It is a figure which shows the difference in structure of 2w9e and 2w9d in amino acid sequence 90-92. It is a figure which shows the difference in structure of 2w9e and 1i4m in the amino acid sequence 151-152. It is a figure which shows the difference in the structure of the whole molecule | numerator of free factor RF2 in 2ihr and 2wh1. It is a figure which shows the difference in the structure of the whole molecule | numerator of ribosomal RNA in 2wrn and 2wh1. In 2wrn, it is a figure which shows the molecular structure which ribosomal RNA and the elongation factor EF-Tu interact. It is a figure which shows the difference in the structure in C366-C372 of the ribosomal RNA in 2wrn and 2wh1. It is a figure which shows the structural pattern of the RNA principal chain characteristic of SARCIN-RICIN LOOP.

Hereinafter, the present invention will be described in detail based on preferred embodiments.
In addition, the description content in the technical literature and patent gazette quoted in this specification shall be referred as description content of this invention.
1. Conformation Homology Evaluation Apparatus of the Present Invention FIG. 1 is a system configuration diagram of a conformation homology evaluation apparatus of the present invention. This conformational homology evaluation apparatus includes a display unit 1, an input unit 2, a processing unit 3, a main storage unit 4, an output unit 5, an external interface (I / O) unit 6, and a database 7. Connected by. Such a conformation homology evaluation apparatus can be constructed by, for example, a personal computer incorporating a program for performing structural homology evaluation.

  The display unit 1 includes a CRT, a liquid crystal display, and the like, and performs screen display for conformational structure homology evaluation. The input unit 2 includes various input means such as a keyboard and is used for inputting data and information required by the user. The processing unit 3 can be composed of a CPU, and performs various controls, calculations, and the like for conformational homology evaluation. The main storage unit 4 stores a program for evaluating structural homology. The output unit 5 includes a printer or the like, and outputs a processing result or the like to the user. An external interface (I / O) unit 6 connects to other terminal devices via a LAN, the Internet, or the like. The database 7 stores various data for conformational homology evaluation in a rewritable manner.

2. Compound to be analyzed in the present invention The conformational homology evaluation apparatus of the present invention basically uses a notation method in which encoding of dihedral angles is defined more strictly than a rule based on the IUPAC nomenclature. Therefore, the compound to be analyzed in the present invention is, in principle, a compound that can be expressed by a chemical structural formula or an amino acid sequence. Any organic compound that can be expressed by a chemical structural formula may have either a chain structure or a cyclic structure as part of the structure. In addition, alkyl groups such as methyl and isopropyl groups, ester groups, carbonyl groups, hydroxyl groups, carboxyl groups, phenyl groups, hydroxyphenyl groups, guanidino groups, thiol groups, selenol groups, imidazolyl groups, amino groups, indolyl groups, alkenes. And may be substituted with a functional group such as amide, disulfide, halogen, oxygen atom, nitrogen atom, phosphorus atom or sulfur atom. Further, it may have a structure that is not a covalent bond, for example, a hydrogen bond or a coordinate bond to a metal ion. The present invention can be applied not only to oligopeptides having about 2 to 10 amino acids linked to them, but also to polypeptides and modified proteins as well as polypeptides having about 10 to 100 bound amino acids, and can be applied to such cases. The constituent amino acids of glycine (Gly), alanine (Ala), valine (Val), leucine (Leu), isoleucine (Ile), phenylalanine (Phe), tyrosine (Tyr), tryptophan (Trp), serine (Ser), Threonine (Thr), Cysteine (Cys), Cystine (CyS-SCy), Methionine (Met), Aspartic acid (Asp), Asparagine (Asn), Glutamic acid (Glu), Glutamine (Gln), Lysine (Lys), Arginine ( Arg), histidine (His), proline (Pro), selenocysteine (Sec), hydroxyproline, thyroxine, O-phosphoserine, β-alanine, etc., any amino acid may be peptide-bonded, Constituent amino acids Les may be further modified by other functional groups. Furthermore, the present invention can be applied to ribonucleic acid (RNA) and modified RNA to which about 2 to 3000 nucleotides are bound as long as they can be expressed by a nucleobase sequence. Any nucleobase such as A), guanine (G), uracil (U), cytosine (C) may be bound to the pentose, and each constituent nucleotide may be further modified with another functional group. . It can also be applied to deoxyribonucleic acid (DNA).
In addition, the present invention extracts the encoded conformation notation and amino acid sequence pattern having high structural homology but different amino acid sequences by combining the encoded conformation notations corresponding to the amino acid sequences. In some cases, the target compounds are limited to peptide-binding compounds such as the above peptides and proteins.

As such an organic compound, the following compounds can be illustrated, for example.
Typical proteins include protein hormones such as insulin, cytokines, various enzymes such as glutathione-S-transferase, glutathione reductase, cyclooxygenase, various receptors such as immunoglobulins or functional fragments thereof, G protein-coupled receptors, etc. In addition to ion channels and transporters, chaperonins, molecular chaperones, human serum albumin, β amyloid protein, prion protein, rhodopsin, G protein and the like can be mentioned. Typical RNAs include messenger RNA (mRNA), transfer RNA (tRNA), ribosomal RNA (rRNA), microRNA, mRNA-like non-coding RNA, small nuclear RNA, small nucleolar RNA, signal recognition particle RNA, etc. It is done. Typical organic compounds other than protein and RNA include glutathione, oxidized glutathione, ibuprofen, ibuprofen dimer, 2-phenylpropionic acid, levofloxacin, piperazine, thalidomide, thalidomide dimer, 5'-hydroxythalidomide, phthalimide , Dioxopiperidine, paclitaxel, paclitaxel tail, paclitaxel tail methyl ester, baccatin III, benzamide, malathion, diethyl succinate, 2-mercaptodiethyl succinate, 2-butanol, 2-pentanol, 2-hexanol, 2-heptanol, 2-octanol, 2-nonanol, 2-decanol, 2-methyl-1-butanol, 3-methyl-1-pentanol, 4-methyl-1-hexanol, 5-methyl-1-heptanol, 6-methyl-1 -Octanol, cis-permethrin, cis-3- (2,2-dichlorovinyl) -2,2-dimethylcyclopropanecarboxylic acid benzyl ester, cis-3- (2,2-dichlorovinyl) -2,2-dimethylcyclopropanecarboxylic acid methyl ester, 3-phenoxybenzyl Alcohol, cholesterol acetate, cholesterol propionate, n-butylic acid cholesterol ester, cholesterol n-valerate, cholesterol n-hexanoate, cholesterol n-heptanoate, cholesterol n-caprylate, cholesterol pelargonate, cholesterol n-caprate, cholesterol laurate Rate, cholesterol myristate, cholesterol palmitate, cholesterol, β-cholestanol, cholesteryl chloride, cholesteryl bromide, cholesterol methyl carbonate, cholesterol ester Rucarbonate, cholesterol-n-butyl carbonate, cholesterol-n-amyl carbonate, cholesterol-n-hexyl carbonate, cholesterol-n-heptyl carbonate, cholesterol-n-nonyl carbonate, cholesterol oleyl carbonate, adenine nucleoside, Guanine nucleoside, uracil nucleoside, cytosine nucleoside, thymine nucleoside, adenosine 5'-phosphate, guanosine 5'-phosphate, uridine 5'-phosphate, cytidine 5'-phosphate, deoxyadenosine monophosphate, deoxyguanosine monolin Examples include acids, deoxythymidine monophosphate, and deoxycytidine monophosphate.

3. Outline of Conformational Homology Evaluation An outline of conformational homology evaluation by the conformation homology evaluation apparatus of this embodiment will be described below.
First, the processing unit 3 receives an input of a molecular model of an analysis target compound.
Next, the processing unit 3 is based on a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts based on the received molecular model. With respect to the structure in which the conformation can be uniquely determined by the one encoded conformation notation, the notation of the chemical bonding site of interest required for the conformation notation is as follows. An encoded conformation notation that is left unnecessarily omitted is extracted from the molecular model, and the extracted encoded conformation notation is stored in the database 7.
Here, extracting the encoded conformation notation that is necessary for the conformation notation and omits unnecessary notation leaving the notation of the chemical bonding site of interest is, for example, in immunoglobulins, When evaluating the structural homology of the hypervariable region of the light chain, only the encoded conformational representation of the corresponding hypervariable region of the light chain is extracted, but the homology of the entire protein molecule including the light chain and heavy chain is extracted. When sex evaluation is performed, this means extracting the encoded conformational notation of the entire immunoglobulin molecule.

  At this time, the conformational structure of the compound to be analyzed is the structural data obtained from the X-ray crystal structure analysis, and it is necessary to compare the structural homology with the structural data that does not include the structural information of the coordinate position of the hydrogen atom. In some cases, for the molecular model that has received the input, a code that can uniquely determine the conformation in accordance with a predetermined priority rule in the dihedral angle classification code involving a hydrogen atom whose coordinate position is not determined. The extracted conformational notation is extracted from the molecular model, and the extracted encoded conformational notation is stored in the database 7.

  Next, the processing unit 3 extracts a plurality of the encoded conformational notations for performing conformational homology evaluation from the database 7 and codes the encoded conformational notations into a form in which homology comparison is possible. Sort and align the format. Subsequently, with respect to a certain dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining dihedral angle classification codes are It is determined that the homology is low, arithmetic processing is performed, and the conformation homology evaluation result is stored in the database 7. More specifically, a set of the dihedral angle classification codes for comparing structural homology is created for one dihedral angle classification code at a corresponding dihedral angle position, and the elements of the set are 360 degrees. Is included in three consecutive dihedral angle classification codes among dihedral angle classification codes determined by dividing into 12 and at least one element of the set is consecutively adjacent If the condition of the dihedral angle classification code in the middle of the three dihedral angle classification codes is satisfied, the homology is determined to be high, and the other is low and the calculation process is performed. The sex evaluation result is stored in the database 7. Sequentially, the dihedral angle position is changed to create a set of dihedral angle classification codes, the same calculation process is performed, and the conformation homology evaluation result is stored in the database 7. At this time, if necessary, only the dihedral angle classification code determined by dividing 360 degrees into 12 is highly homologous in a format rearranged into a form that allows comparison of homology. The dihedral angle classification code is determined to have low homology, and calculation processing is performed, and the conformation homology evaluation result is stored in the database 7. The processing unit 3 takes out the homology evaluation result from the database 7 together with the encoded conformation notation and displays it on the display unit 1 and outputs it to the output unit 5 or an external interface (I / O) unit 6 as necessary. Data transfer is performed via

  When the analysis target compound to be compared is a protein or a peptide, the processing unit 3 has a corresponding relationship with the amino acid sequence in a form in which rearrangement is performed so that the homology can be compared as necessary. In combination with the encoded conformation notation, an arithmetic process for extracting an encoded conformation notation and amino acid sequence pattern having high structural homology but different amino acid sequences is performed, and the processing result is stored in the database 7 Displayed on the display unit 1. The processing unit 3 outputs data to the output unit 5 or transfers data via the external interface (I / O) unit 6 as necessary.

A detailed description will be given below for each step according to the flowchart of FIG.
3-1. Input process of molecular model (S1 and 2)
In the present invention, first, a molecular model is input (steps S1 and S2). The input molecular model is obtained from X-ray crystal structure analysis data, data optimized by theoretical calculation such as density functional calculation (Non-patent Documents 1-5), or infrared circular dichroism spectrum. Any molecular structure data of the experimental value data (Non-patent Documents 1-5) may be used. In this case, for example, a compound name, a chemical structural formula, and a molecular model are associated in advance and the information is stored in the database 7. When the user inputs a compound name or chemical structural formula from the input unit 2, the corresponding molecular model is displayed. You may be able to convert. In order to input the molecular model of the conformation structure of each of a plurality of analysis target compounds, for example, it is realized by selecting the molecular structure data file using the MolWorks software manufactured by Beyond Computing and opening the file. I can do it.

3-2. Code conversion to code conformation notation and code rearrangement (S3-6)
Next, when comparing structural homology between data that also includes information on the bonded hydrogen atoms, such as molecular models that have been optimized by theoretical calculations such as density functional theory calculations, Code conversion to the encoded conformational notation described in Patent Document 4 is performed (steps S3 to S4). In addition, when X-ray crystal structure data is used as a molecular model to be compared, and hydrogen atoms are selected such that hydrogen atoms are selected when the priority rules are followed, the X-ray crystal structure data Since there is no information regarding hydrogen atoms, the process proceeds to step S5.
The encoded conformational notation consists of a combination of a “dihedral angle position code” corresponding to each chemical bonding site and a “dihedral angle classification code” determined by dividing 360 degrees into 12 parts. When the “corner classification code” is determined by the priority rule, it is used by substituting the code of the “dihedral angle classification code” into the specified code in the “dihedral angle position code” (Patent Literature). 4).
Here, the “priority rule for determining the dihedral angle classification code” conforms to the IUPAC nomenclature. In other words, the IUPAC nomenclature specifies that the conformation is expressed based on the dihedral angle formed by the ligand (substituent or atom) bound to each end of the chemical bond of interest. As a rule,
(1) When all of the ligands are different, the one with the most precedence according to the ranking rule determined by the R / S display,
(2) When only one is different from the others,
(3) When all are the same, priority is given to the one with the smallest dihedral angle.
In order to determine the dihedral angle classification code, a substituent or an atom is selected according to the above priority rule, and the angle of the dihedral angle formed by the substituent or atom is determined. The code corresponding to the dihedral angle is selected according to “Method”. For conversion from dihedral angle information to the dihedral angle classification code, a coded conformational notation prototype application (H21PRO-1028) owned by the National Institute of Advanced Industrial Science and Technology (H21PRO-1028) (non-patent document) 1) can be used.

<Application example of dihedral angle position code>
As a specific example, as described in Patent Document 4, when there is a structure in which the conformation can be uniquely determined by one conformation notation such as when having a ring structure, notation of the chemical bonding site to be noted The unnecessary notation is omitted, and the code conversion to the encoded conformation notation that uniquely indicates the conformation is performed (step S4). If it is more convenient to separate the compounds into fragments from the viewpoint of creating a database, a dihedral angle position code combining a prefix of the fragment and a code indicating the dihedral angle position of the chemical bonding site can be used. In the case of protein molecules, such a notation is suitable. For example, as shown in FIG. 3, a prefix corresponding to the amino acid sequence and dihedral angles required for the main chain and side chain of each amino acid. A dihedral angle position code can be expressed by combining the codes indicating the positions. In that case, it is necessary to indicate the dihedral angle position where detailed position information of the hydrogen atom is involved in the determination of the dihedral angle classification by the priority rule, and the position of the hydrogen atom such as X-ray crystal structure analysis data cannot be determined When using data, since the sign of a predetermined dihedral angle classification code based on the determination of the dihedral angle classification based on the priority rule cannot be used, only the corresponding dihedral angle position is used instead. The sign of the meaning dihedral angle position code is used.

As an example, a procedure for expressing the amino acid sequence of the following insulin A chain as a character string is shown.
“GLY ILE VAL GLU GLN CYS CYS THR SER ILE CYS SER LEU TYR GLN LEU GLU ASN TYR CYS ASN” (SEQ ID NO: 1)
The dihedral position code of insulin A chain is the prefix corresponding to the amino acid sequence and the dihedral position of the main chain (A, B, C) and side chain (a, b, c, d, etc.) of each amino acid. The following can be described using a combination of symbols indicating:
"GLY-ABC-ILE-A-ab-BC-VAL-Aa-BC-GLU-A-abcd-BC-GLN-A-abcd-BC-CYS-A-ab (c) -BC-CYS-A- ab-BC-THR-A-ab-BC-SER-A-ab-BC-ILE-A-ab-BC-CYS-A-ab (c) -BC-SER-A-ab-BC-LEU-A -ab-BC-TYR-A-abc-BC-GLN-A-abcd-BC-LEU-A-ab-BC-GLU-A-abcd-BC-ASN-A-abc-BC-TYR-A-abc -BC-CYS-A-ab-BC-ASN-A-abc-B "
By using the method described below, the conformational structure is uniquely described as a character string by substituting the dihedral angle classification code for the dihedral angle position of each amino acid main chain and side chain. can do.

<Example of dihedral angle classification code>
Here, an example of a method for attaching a dihedral angle classification code determined by dividing 360 degrees into 12 parts will be described. Usually, as described in Patent Document 4, 360 degrees is divided into 6 parts, and a predetermined code such as 1-6 shown in FIG. Indicate the angle classification. In rare cases, the conformation that should be judged as two different conformations in this classification may be expressed as the same code. In that case, each of the six divided parts is further divided into two parts. Then, a predetermined code such as α (a, clockwise) and β (b, counterclockwise) is attached to the divided part, and a combination such as these two types of codes is mainly used. The dihedral angle classification is described using symbols corresponding to the 12 divided parts shown in FIG. Furthermore, the code shown in FIG. 4 may be used on the boundary line in case the corresponding dihedral angle is accidentally located on the boundary line.

<Signs of σ (cis) and τ (transformer)>
Furthermore, as described in Patent Document 4, in the rules based on the IUPAC nomenclature, (3) when all divertants or atoms have the same dihedral angle is selected, one code In this case, a new code such as σ (s, cis) or τ (t, trans) based on the relative positional relationship may be generated. Use it to be able to express the conformation uniquely. For example, in (S) -ibuprofen, when determining the dihedral angle in the aromatic ring to which the propionic acid moiety is bonded, (3) when all the substituents or atoms are the same, the dihedral angle that is the smallest is selected. In this case, in the encoded conformation notation ibut-3α2α (phpa-3α2α), there are two types of conformational structures represented by the same notation shown in FIG. In that case, the relative positional relationship between the bonds of interest is used to distinguish using the signs of σ (cis) and τ (transformer).

  Further, as described in Patent Document 4, in the rules based on the IUPAC nomenclature, (3) when all the substituents or atoms are the same, the dihedral angle that is the smallest is selected. It can happen that the angle formed by the dihedral angles of substituents or atoms that are indistinguishable in the clockwise direction is the same. In that case, since a plurality of encoded conformational representations are possible in one molecular model, the one with the smallest dihedral angle in the clockwise direction is prioritized. For example, in the dihedral angle of the piperazine ring shown in FIG. 6 in levofloxacin, since the piperazine ring itself is symmetrical, the angle formed by the dihedral angle of the substituents is positive and negative and the absolute value is the same. Encoded conformational notation can be taken. By applying the priority rule, the conformational notation is uniquely defined as levo-2β-1β5α (pipa-1β1β).

In the conformation homology evaluation apparatus of the present embodiment, the conformation structure of the analysis target compound is structure data obtained from the X-ray crystal structure analysis, and the structure does not include the structure information of the coordinate position of the hydrogen atom. When it is necessary to compare structural homology with the data, the conformation is determined according to the priority rule defined in advance in the dihedral angle classification code involving the hydrogen atom whose coordinate position is not determined. An encoded conformation notation that can uniquely determine the locus is extracted from the molecular model, and code conversion is performed (steps S3 to S5). That is, instead of the encoding rules in step S4, code conversion to encoded conformational notation is performed according to the following priority rules (step S5). Even if the priority rule is applied but cannot be determined, the comparison may be made only with the corresponding dihedral angle positions. In order to show that, what cannot be determined is described using a code such as “X”.
<Priority rules for coding when the position of the hydrogen atom is not fixed>
(A) In accordance with the IUPAC nomenclature rules regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the bond plane is the smallest in the clockwise direction when viewed from the hydrogen atom.
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.

In the case of the above-described insulin A chain, by substituting the dihedral angle classification code converted from the X-ray crystal structure data “2c8r” of insulin sequentially registered in PDBj into the dihedral angle position code. ,
`` GLY-X4α4α-ILE-3α-1α1β-5α4β-VAL-3β-3β-5α4β-GLU-3β-3β2α2βX-5α4α-GLN-3β-3α3α5αX-1β4α-CYS-5β-3β6α (2α) -1β4α-CYS- 5α-1βX-5α4β-THR-3β-2βX-1β4β-SER-6β-2αX-4β4α-ILE-3β-3β1β-3α4β-CYS-1β-1α2β (2α) -4β4β-SER-6β-2αX-4β4β-LEU -3β-1β1α-5α4β-TYR-3α-1α3βX-5α4α-GLN-3β-3β1α5βX-5α4β-LEU-3α-3β3β-5α4α-GLU-3β-3β1α3αX-1β4β-ASN-3β-3β1βX-1β4β-TYR-6α -3β3βX-1β4β-CYS-1β-3αX-3α4β-ASN-3β-3β3αX-4β ''
The conformation structure can be uniquely described as a character string. Here, even in the encoded conformational notation for X-ray crystal structure data, together with information such as physical property values required after analysis in association with the conformational notation according to the method described in Patent Document 4 The encoded conformational notation may be displayed. That is, for the molecular model corresponding to the encoded conformational notation, the optimum is obtained from comparison with the measured physical property value such as analysis of infrared circular dichroism spectrum and theoretical calculation such as density functional calculation. You may compare the homology of the physical property value of a chemical structure.

  Next, a format in which a plurality of stored encoded conformational notation data can be retrieved from the database, data is aligned according to the correspondence relationship of the dihedral angle position codes, and the homology of the dihedral angle classification codes can be compared Rearranging is performed (step S6). The rearrangement may be performed for each fragment, for each dihedral angle position code, for each combination of a plurality of fragments or dihedral angle position codes. According to the method described in Patent Document 4, along with the encoded conformational notation in which rearrangement is performed, data of the analysis target compound related to the conformation such as the measured physical property value or the predicted value from the theoretical calculation is accompanied. May be.

3-3. Conformational homology evaluation by “swing tolerance calculation analysis method” Furthermore, in the conformational homology evaluation apparatus of this embodiment, there is a correspondence relationship in the encoded conformation notation that has been rearranged. The homology evaluation of the dihedral angle classification code is performed (steps S7 to S9). Many of the ligand molecules having physiological activity have a plurality of stable conformational structures in the solution state, and the equilibrium state is successfully utilized for the interaction with the receptor (Non-patent Documents 4 and 5). Each conformational notation of a molecule should have features that can tolerate structural fluctuations that are also associated with polymorphs in its crystal structure. Therefore, in the conformation homology evaluation apparatus of the present embodiment, in the dihedral angle classification code in the correspondence relationship, the dihedral angle classification code itself adjacent to the dihedral angle classification code so as to cope with the problem of fluctuation of the molecular structure. The combination of angle classification codes has high homology, and the remaining dihedral angle classification codes are determined to have low homology, and calculation processing is performed to evaluate homology (step S7).

That is, for one dihedral angle classification code in a corresponding dihedral angle position, a set of the dihedral angle classification codes for structural homology evaluation is created, and the elements of the set divide 360 degrees into 12 parts. Are included in three consecutive dihedral angle classification codes among the dihedral angle classification codes determined in advance, and at least one element of the set is the three consecutive adjacent dihedral angle classification codes. If the dihedral angle classification code satisfies the condition of the central dihedral angle classification code, it is judged that the homology is high and the homology is low otherwise, and the homology is evaluated. (Step S7). In step S7, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code at each position in the corresponding dihedral angle positions, and the elements are continuously adjacent as elements. Create a set with three dihedral angle classification codes as elements, compare the elements of the set at the position in each analysis target compound, and at least one element among the elements of the set belonging to each compound Search whether or not the condition is common to all compounds, and the common element is a central element in the set belonging to at least one compound,
It can be realized by performing arithmetic processing for determining that the homology is high when the condition is satisfied, and that the homology is low otherwise, and evaluating the homology.

More specifically, when the dihedral angle classification is described using the codes corresponding to the 12-divided parts shown in FIG. 4, a set of three adjacent dihedral angle classification codes is obtained. X
X = {4α, 2β, 2α}, {2β, 2α, 5β}, {2α, 5β, 5α}, {5β, 5α, 1β}, {5α, 1β, 1α}, {1β, 1α, 6β}, {1α, 6β, 6α}, {6β, 6α, 3β}, {6α, 3β, 3α}, {3β, 3α, 4β}, {3α, 4β, 4α}, {4β, 4α, 2β}
It is expressed. Here, we simplify one combination of set X,
Y = {p, q, r}, Y⊂X
It will be expressed as With respect to one dihedral angle classification code at a corresponding dihedral angle position, a set of the dihedral angle classification codes for homology evaluation is created, and elements of the set are determined by dividing 360 degrees into 12 parts. The dihedral angle classification code is always included in three consecutive dihedral angle classification codes, and at least one element of the set is the three adjacent dihedral surfaces In order to determine that the homology is high when the condition of the central dihedral angle classification code among the angle classification codes is satisfied, and other than that, the homology is low. Assuming that the elements of the set are x, all x are related to a single combination Y of the set X.
x∈Y
And at least one of x is
x = q
It is only necessary to perform arithmetic processing that determines that the homology is high when the condition is satisfied and that the homology is low in all other cases. If the corresponding dihedral angle is coincidentally located on the 4β, 4α boundary line, 5α if the corresponding dihedral angle is located on the 4α, 2β, 2α boundary line. 1β, 1β, 1α, if located on the boundary of 1β, 6β, 6α, if located on the boundary of 6β, 3β, 3α That's fine.

  Furthermore, in the encoded conformation notation that has been rearranged, the dihedral angle position code is changed, and the dihedral angle position code in one place or the combination of the dihedral angle position codes in a plurality of places is subjected to homology evaluation. A set of dihedral angle classification codes is created, and arithmetic processing on the same dihedral angle classification codes is sequentially performed to evaluate structural homology. When evaluating structural homology for a combination of the dihedral angle position codes at a plurality of locations, in each of the dihedral angle position codes, after performing homology calculation processing on the dihedral angle classification code, What is necessary is just to perform the logical operation of the logical product (AND) of the combination of the said dihedral angle position code of several places. At this time, using the structural homology evaluation result, the ratio of high homology may be calculated.

  In addition, when verifying fine differences in the molecular structure of crystal polymorphs, only the dihedral angle classification code itself determined by dividing 360 degrees into 12 parts is highly homologous as necessary. It is determined that the face angle classification code has low homology, arithmetic processing is performed, and homology is evaluated (step S9). At this time, the ratio of high homology may be calculated at the same time.

  Furthermore, in the case of structural homology evaluation of interaction sites between ligand molecules and protein molecules with structural changes, it is necessary to deal with structural fluctuations and compare side chain structural homology as necessary. 1) A combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous (○). 2) One dihedral angle classification code. The homology of the combination of dihedral angle classification codes that are relatively close to each other between the codes is moderate (△). 3) The homology of the combination of other dihedral angle classification codes is determined to be low (×), and the processing is performed. Structural homology evaluation may be performed.

  That is, for the one dihedral angle classification code at the corresponding dihedral angle position, a set of the dihedral angle classification codes for performing structural homology evaluation is created. 1) The element of the set is set to 360 degrees. Of the dihedral angle classification codes determined by division, the dihedral angle classification codes are always included in three consecutive dihedral angle classification codes, and at least one element of the set is adjacent to the consecutive The homology is high when the condition of the central dihedral angle classification code among the three dihedral angle classification codes is satisfied (◯), 2) the elements of the set are determined by dividing 360 degrees into 12 parts Three dihedral angle classification codes that are always included in the three adjacent dihedral angle classification codes among the dihedral angle classification codes, and that all elements of the set are adjacent to each other. When the condition that is not the central dihedral angle classification code is satisfied , Moderate homology (△), otherwise low homology (×) and determined by the arithmetic processing may be subjected to structural homology evaluation.

More specifically, when the dihedral angle classification is described using the codes corresponding to the 12-divided parts shown in FIG. 4, a set of three adjacent dihedral angle classification codes is obtained. X
X = {4α, 2β, 2α}, {2β, 2α, 5β}, {2α, 5β, 5α}, {5β, 5α, 1β}, {5α, 1β, 1α}, {1β, 1α, 6β}, {1α, 6β, 6α}, {6β, 6α, 3β}, {6α, 3β, 3α}, {3β, 3α, 4β}, {3α, 4β, 4α}, {4β, 4α, 2β}
It is expressed. Here, we simplify one combination of set X,
Y = {p, q, r}, Y⊂X
It will be expressed as For the one dihedral angle classification code in the corresponding dihedral angle position, a set of the dihedral angle classification codes for structural homology evaluation is created, and 1) the elements of the set divide 360 degrees into 12 parts Are included in three consecutive dihedral angle classification codes among the dihedral angle classification codes determined in advance, and at least one element of the set is the three consecutive adjacent dihedral angle classification codes. If the condition of the central dihedral angle classification code among the dihedral angle classification codes is satisfied, the homology is high (◯). 2) The two elements whose elements of the set are determined by dividing 360 degrees into 12 parts Of the three dihedral angle classification codes that are always included in the three adjacent dihedral angle classification codes of the angle classification code, and that all elements of the set are among the three adjacent dihedral angle classification codes Homology if a condition that is not a central dihedral angle classification code is met To determine medium (△) and low homology (x) otherwise, let x be an element of the set of dihedral angle classification codes for homology evaluation, and all x are In relation to a single combination Y,
x∈Y
And at least one of x is
x = q
High homology when satisfying (○),
Every x is related to a single combination Y in the set X.
x∈Y
And all x
x ≠ q
The homology is moderate (△) when
In all other cases, an arithmetic process for determining that the homology is low (x) may be performed. If the corresponding dihedral angle is accidentally located on the 4β, 4α boundary line, the corresponding dihedral angle is located on the 4α, 2β, 2α boundary line, and if the corresponding dihedral angle is located on the 2α, 5β, 5α boundary line, 5α , 1β, 1α, if it is located on the boundary of 1β, 6β, 6α, if it is located on the boundary of 6β, 3β, 3α That's fine.

3-4. As an amino acid sequence, such as when determining whether or not a protein molecule has undergone a structural change when applying a “swing tolerance calculation analysis method” when a conformational structure can be selected as a reference in the compounds to be compared If there is no change, it is only necessary to perform a homology comparison of only the conformational structure.In other words, the encoded conformation notation of the selected specific protein molecule such as the protein molecule before the change is used as another reference value. A homologous comparison with the encoded conformational representation of one or more protein molecules may be performed. In that case, for example, the conformational structure of the protein molecule before the change is always selected as a reference, so that the dihedral angle classification codes of the conformational structure are all three consecutively adjacent. The dihedral angle classification code at the center. On the other hand, all elements of the set of dihedral angle classification codes at the corresponding dihedral angle positions of the protein molecules for which structural homology evaluation is performed are adjacent to the median value of the reference dihedral angle classification code or adjacent to it. If it matches with the dihedral angle classification code, the homology is high, and other than that, it is determined that the homology is low.

Specifically, the dihedral angle classification code for evaluating the structural homology at the corresponding dihedral angle position selected from the encoded conformational notation of a specific protein molecule as v is defined as v. Assume that dihedral angle classification codes adjacent to each other are arranged in the order of a set of Z = {u, v, w}. When a set of all three consecutively adjacent dihedral angle classification codes is X, and the dihedral angle classification is expressed using the codes corresponding to the 12 divided parts shown in FIG.
X = {4α, 2β, 2α}, {2β, 2α, 5β}, {2α, 5β, 5α}, {5β, 5α, 1β}, {5α, 1β, 1α}, {1β, 1α, 6β}, {1α, 6β, 6α}, {6β, 6α, 3β}, {6α, 3β, 3α}, {3β, 3α, 4β}, {3α, 4β, 4α}, {4β, 4α, 2β}
And
Z⊂X
Z can be expressed as any one notation in the set X. With respect to one dihedral angle classification code at a corresponding dihedral angle position, the dihedral angle classification code of the specific protein molecule is always selected as a reference, and the structural homology evaluation at the corresponding dihedral angle position is performed. If all elements of the set of classification codes match the standard dihedral angle classification code or the dihedral angle classification code adjacent to the dihedral angle classification code, the homology is high. In order to determine low, a set of the dihedral angle classification codes for performing homology evaluation at a corresponding dihedral angle position including the selected dihedral angle classification code of the specific protein molecule is created, and its elements And y, all y
y = u
Or
y = v
Or
y = w
Processing that determines that the homology is high when the above condition is satisfied and that the homology is low in all other cases may be performed. If the corresponding dihedral angle is coincidentally located on the 4β, 4α boundary line, 5α if the corresponding dihedral angle is located on the 4α, 2β, 2α boundary line. 1β, 1β, 1α, if located on the boundary of 1β, 6β, 6α, if located on the boundary of 6β, 3β, 3α That's fine.

3-5. Extraction of amino acid sequence patterns with encoded conformational notation corresponding to secondary structure
It is well known that α-helix, β-sheet and the like, which are well-known secondary structures of proteins and peptides, form the same pattern even if the amino acid sequences are completely different.
By using the “sway tolerance calculation analysis method” of the present invention, the same secondary structure is taken, including structures other than α helix and β sheet such as random coil, within the permissible range of shake, although the amino acid sequences are different. Encoded data (amino acid sequence + structure information) and semantic information can be extracted. In a useful protein or peptide, in order to modify the amino acid sequence without changing the three-dimensional structure of the active domain, it is possible to obtain encoded data that is a candidate for amino acid replacement. Collecting these encoded data and creating a database can be used particularly for improving various enzyme activities, producing humanized antibodies (CDR antibodies), chimeric antibodies, and the like.
Therefore, in the conformational homology evaluation apparatus of this embodiment, when the analysis target compound to be compared is a protein or peptide, an amino acid sequence having high structural homology in consideration of fluctuation of the molecular structure, if necessary. Extract the pattern, extract the ones that do not match the amino acid sequence, and extract the amino acid sequence pattern with encoded conformation notation corresponding to other secondary structures such as α helix, β sheet, random coil, etc. (Steps S10 to S11). At this time, according to the method described in Patent Document 4, along with the encoded conformation notation, data such as measured physical property values and predicted values from theoretical calculations related to the extracted amino acid sequence pattern, or homology analysis results May be accompanied.

3-6. Final step of conformational homology evaluation (S12 and 13)
Finally, after performing the steps S1 to S11 as many times as necessary (step S12), display and / or output of necessary data such as the homology evaluation result (step S13).

  In addition, the process from step S7 to S9 may be performed before and after. Further, for example, according to the method described in Patent Document 4, a conformation structure in which a protein and a ligand molecule interact with each other, such as a conformation structure of a bioactive ligand molecule incorporated into a protein, is subjected to the above-described process. Conformational homology evaluation may be performed.

3-7. Structural pattern analysis by the improved “swing tolerance calculation analysis method” In conformational homology evaluation up to 3-6, even if structural homology evaluation can be performed, secondary structure itself such as α helix, β sheet, random coil, etc. The fragment structure information cannot be extracted from the code itself because of the fluctuation of the molecular structure.
To extract secondary structure such as α helix, β sheet, and random coil in protein fragment and stem structure and loop structure in RNA fragment, the following may be performed.

First, for molecular models that are known to have secondary structures such as α-helix, β-sheet, random coil, etc., stem structure, loop structure, etc. to be analyzed, only fragments corresponding to the characteristic structural region Then, a molecular model is input according to the process of 3-1, and subsequently, code conversion to code conformation notation and code rearrangement are performed according to the process of 3-2.
Next, the dihedral angle classification code having the highest appearance frequency in each dihedral angle position code constituting the fragment is extracted, and the extracted encoded conformational notation is expressed as α helix, β sheet, random coil, stem The data is stored in the storage unit as a template of a structure pattern such as a structure or a loop structure.

  Separately, with respect to the molecular model of the conformational structure of the analysis target compound, the molecular model is input according to the step 3-1, and subsequently the code conversion and the code rearrangement to the encoded conformation notation are performed according to the step 3-2. I do.

3-5. In the case of the method for extracting an amino acid sequence pattern described in 3), basically, the above 3-4. Because of the application of the “sway tolerance calculation analysis method”, amino acid sequences and nucleic acid base sequences that match the structural patterns of α helix, β sheet, random coil, stem structure, loop structure, etc., which are more strictly defined The main focus is on extracting information, and this method is suitable for such cases.
On the other hand, in this section, a compound to be analyzed consisting of a very large number of fragments such as amino acid fragments and nucleotide fragments, an α helix in a broader sense, β sheet, random coil, Since the main purpose is to evaluate and analyze whether or not it has a known characteristic structure pattern common to the stem structure, loop structure, etc., the above 3-4. The following calculation process is performed by improving a part of the “sway allowable calculation analysis method”.

That is, the encoded conformation notation that matches each structural pattern and the encoded conformation notation of the analysis target compound are extracted from the storage unit, and the encoded conformation notation that matches each structural pattern is compared. As a reference for the subject,
At each position in the corresponding dihedral angle position, the dihedral angle classification code on both sides is extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Compare each analysis target compound and the elements of the set at the position in the reference structure pattern,
Search for whether or not the condition that at least one element among the elements of the set belonging to each fragment is common to all fragments is satisfied,
When satisfying, an arithmetic processing is performed to determine as a fragment that matches each structural pattern, and when not satisfying, it is determined as a fragment having another structural pattern.

As a specific analysis method, the following procedure is followed.
The processing unit accepts an input of a molecular model of a conformation structure of a fragment corresponding to one or more reference structural patterns,
The processing unit describes the dihedral angle position code corresponding to the chemical bonding site of interest for the molecular model that has received the input, and 360 degrees is divided into 12 for each dihedral angle position code. By applying the encoded conformation notation that combines the specified dihedral angle classification codes, the conformation regarding each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
The extracted encoded conformation notation is stored in advance in the storage unit as a reference structure pattern,
Separately, accepting the input of the molecular model of the conformational structure of the compound to be analyzed,
In the processing unit, for the molecular model of the analysis target compound that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Next, an extracted conformational notation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for the structural pattern analysis is extracted,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, the encoded conformational notation of the analysis target compound is taken out from the storage unit, and the reference structure pattern to be compared is selected from the reference structure patterns stored in the storage unit in advance. , Taking out the encoded conformational notation that matches the structure pattern as a reference for comparison,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrasting elements of the set at the positions in the structural pattern to be analyzed and the reference structural pattern,
Search for whether or not the condition that at least one element is common to the reference structure pattern among the elements of the set to which each dihedral angle classification code belongs,
If it satisfies, it is determined that it matches the reference structure pattern, and if it does not, it is determined that it corresponds to the other structure pattern.
Subsequently, an operation for performing the same calculation process at the next position is performed, and the operations for all corresponding positions are repeated.

More specifically, when the dihedral angle classification is described using the codes corresponding to the 12-divided parts shown in FIG. 4, a set of three adjacent dihedral angle classification codes is obtained. X
X = {4α, 2β, 2α}, {2β, 2α, 5β}, {2α, 5β, 5α}, {5β, 5α, 1β}, {5α, 1β, 1α}, {1β, 1α, 6β}, {1α, 6β, 6α}, {6β, 6α, 3β}, {6α, 3β, 3α}, {3β, 3α, 4β}, {3α, 4β, 4α}, {4β, 4α, 2β}
It is expressed. Here, we simplify one combination of set X,
Y = {p, q, r}, Y⊂X
It will be expressed as At each position in the corresponding dihedral angle position, the dihedral angle classification code on both sides is extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Compare each analysis target compound and the elements of the set at the position in the reference structure pattern,
Search for whether or not the condition that at least one element among the elements of the set belonging to each fragment is common to all fragments is satisfied,
If it is satisfied, it is determined as a fragment that matches each structure pattern. If it is not satisfied, it is determined as a fragment having other structure patterns.
A set of dihedral angle classification codes is created at the first position in the corresponding dihedral angle position in the fragment for which structural pattern analysis is performed, and the element of the dihedral angle classification code set for structural pattern analysis is x Then, every x is related to one combination Y in the set X,
x∈Y
If the condition is satisfied, an arithmetic process may be performed such that the fragment is matched with the reference structure pattern at the first position, and if not satisfied, the fragment is determined as a fragment having a structure pattern other than the reference. Furthermore, a set of dihedral angle classification codes for a single dihedral angle position code or a combination of dihedral angle position codes at a plurality of locations where the dihedral angle position is changed and structural pattern analysis is performed as necessary. Create and sequentially perform the arithmetic processing on the same dihedral angle classification code, and perform the structural pattern analysis. When structural pattern analysis is performed for a combination of dihedral angle position codes at a plurality of locations, the homology calculation processing for the dihedral angle classification code is performed on each dihedral angle position code, and then a plurality of locations are analyzed. The logical operation of the logical product (AND) of the combinations of the dihedral angle position codes is sufficient. Eventually, when the operation result is obtained in the logical operation of the logical product (AND) of the combinations of all the dihedral angle position codes in the correspondence relationship in the fragment, that is, in the correspondence relationship in the fragment When an operation result indicating that the reference structure pattern matches at all dihedral angle positions is obtained, an operation process for determining as a fragment that completely matches the reference structure pattern may be performed. At this time, the ratio of high homology may be calculated using the analysis result of the structure pattern at the same time. If the corresponding dihedral angle is coincidentally located on the 4β, 4α boundary line, 5α if the corresponding dihedral angle is located on the 4α, 2β, 2α boundary line. 1β, 1β, 1α, if located on the boundary of 1β, 6β, 6α, if located on the boundary of 6β, 3β, 3α That's fine.

  3-7. As described above, the improved “swing tolerance calculation analysis method” is used mainly for comparative analysis with known structural patterns, but can also be applied to the extraction and analysis of new structural patterns. Furthermore, it can also be used for homology evaluation of conformational structures in a plurality of compounds to be analyzed. For the same analysis object, 3-7. The improved “swing tolerance calculation analysis method” shown in FIG. 3 is a homology analysis method based on the conventional method (Patent Document 4), 3-3. "Swing tolerance calculation analysis method" shown in 3-4. Applied in parallel with at least one of the “swing tolerance calculation analysis methods” shown in FIG. 1, and operations such as logical product, logical sum, negative logical product, negative logical sum, exclusive logical sum, etc. are obtained for the obtained calculation results. Processing may be performed.

  Finally, after performing the above operations as many times as necessary, necessary data such as structural pattern analysis results are displayed and / or output. In addition, 3-7. The improved “swing tolerance calculation analysis method” shown in 1), the homology analysis method by the conventional method (Patent Document 4), 3-3. "Swing tolerance calculation analysis method" shown in 3-4. To perform evaluation and analysis using the “Swing Permissible Calculation Analysis Method” shown in, you can use the “Swing Permitted Operation Analysis Template” (H22PRO-1102) owned by the National Institute of Advanced Industrial Science and Technology. .

  EXAMPLES Next, although an Example demonstrates this invention still in detail, this invention is not limited to the following examples.

(Example 1) Application of shake-allowing calculation analysis method to X-ray crystal structure data of insulin A chain and B chain
For the X-ray crystal structure data 2vk0_AB and 2c8r of insulin registered in PDBj, using the method described in Patent Document 4, the dihedral angle position corresponding to each chemical binding site based on the molecular model that accepted the input. Coding conversion was performed on the coded conformational notation composed of a combination of a code and a dihedral angle classification code determined by dividing 360 degrees into 12 parts.
Insulin is composed of 21 amino acid residues A chain and 30 amino acid residues B chain connected via two disulfide bonds. . At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data. The character string of the 2c8r A chain at that time is as follows.
`` GLY-X4α4α-ILE-3α-1α1β-5α4β-VAL-3β-3β-5α4β-GLU-3β-3β2α2βX-5α4α-GLN-3β-3α3α5αX-1β4α-CYS-5β-3β6α (2α) -1β4α-CYS- 5α-1βX-5α4β-THR-3β-2βX-1β4β-SER-6β-2αX-4β4α-ILE-3β-3β1β-3α4β-CYS-1β-1α2β (2α) -4β4β-SER-6β-2αX-4β4β-LEU -3β-1β1α-5α4β-TYR-3α-1α3βX-5α4α-GLN-3β-3β1α5βX-5α4β-LEU-3α-3β3β-5α4α-GLU-3β-3β1α3αX-1β4β-ASN-3β-3β1βX-1β4β-TYR-6α -3β3βX-1β4β-CYS-1β-3αX-3α4β-ASN-3β-3β3αX-4β ''

Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. Despite complete amino acid sequence matching, 2vk0_AB and 2c8r are 60.0% in the structural homology comparison of N-terminal main chain by dihedral angle classification code determined by dividing 360 degrees into 12 parts. (A chain), there is only about 80.0% (B chain) homology, and structural homology comparison of the main chain on the C-terminal side is only about 19.0% (A chain), 52.0% (B chain), There were fluctuations in the peptide structure in the X-ray crystal structure data.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining The dihedral angle classification code was judged to have low homology, and was processed and compared for structural homology. The structural homology comparison of the N-terminal main chain revealed 95.0% (A chain), 100.0% (B Chain) homology, 95.2% (A chain), 96.0% (B chain) homology, 90.5% (A chain), 96.2% (main chain as a whole) (B-chain) homology, and it was found that homology comparison can be performed in consideration of structural fluctuations. Table 1 shows the encoded conformational notation and homology analysis results of insulin (2vk0_AB, 2c8r) A chain, and Table 2 shows the encoded conformation notation and homology analysis results of insulin (2vk0_AB, 2c8r) B chain. Shown in (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)
In addition, if the amino acid sequence corresponding to the α helix and β sheet enclosed in the square is processed in consideration of the structure fluctuation, each secondary structure can be expressed in the encoded conformation notation. For example, in the N-terminal main chain (dihedral angle position code A), the sign {6α, 3β, 3α} is an exception because only the bond selected from the priority rule is different for CYS, {5β, 5α, 1β} In the main chain on the C-terminal side (dihedral angle position code BC), centering on the notation of one of the symbols {5β4β, 5β4α, 5α4β, 5α4α, 1β4β, 1β4α} In the case of a β sheet, the N-terminal main chain (dihedral angle position code A) is an exception because the sign {1α, 6β, 6α, 3β} (only the bond selected from the priority rule for CYS is different, { 2α, 5β, 5α}) and the main chain on the C-terminal side (dihedral angle position code BC In, {3β4β, 3β4α, 3α4β, 3α4α, 4β4β, 4β4α} is constructed around a one notation of the code of the secondary structure of the protein was found to be pattern recognition. Furthermore, it was extracted that there was a slight structural difference only with the 11th CYS of the A chain at the binding position with the B chain.
This clearly shows that there are crystal polymorphs due to the conformation capable of expressing the same function in insulin.

[Table 1] Coding conformation notation and analysis results of insulin (2vk0_AB, 2c8r) A chain

[Table 2] Coding conformation notation and analysis result of insulin (2vk0_AB, 2c8r) B chain

(Example 2) Presence of secondary structure difference in immunoglobulin light chain having high amino acid sequence homology
The X-ray crystal structure data 2imm and 2mcp of immunoglobulin light chain (MUS MUSCULUS, HOUSE MOUSE) with high homology of amino acid sequences registered in PDBj was accepted using the method described in Patent Document 4 Based on the molecular model, coding conversion to the coded conformational notation consisting of a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts Went. A molecular model of the X-ray crystal structure data 2imm is shown in FIG. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. Despite the amino acid sequence homology of 99.1%, 2imm and 2mcp are about 54.9% in the structural homology comparison of N-terminal main chain by dihedral angle classification code determined by dividing 360 degrees into 12 In comparison of the structural homology of the C-terminal main chain, the homology was only about 27.4%, and the protein structure was shaken in the X-ray crystal structure data.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining When the dihedral angle classification code was judged to have low homology, it was processed and the structural homology was compared. The structural homology comparison of the N-terminal main chain showed 92.0% homology and the C-terminal main homology. In the structural homology comparison of the chains, the homology was 95.6%, and the entire main chain was 89.5%, indicating that the homology can be compared in consideration of the structural fluctuation. Table 3 shows the encoded conformation and homology analysis results of the immunoglobulin light chain (2imm, 2mcp). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)
In addition, if the amino acid sequence corresponding to the α helix and β sheet enclosed in the square is processed in consideration of the structure fluctuation, each secondary structure can be expressed in the encoded conformation notation. For example, in the N-terminal main chain (dihedral angle position code A), the sign {6α, 3β, 3α} is an exception because only the bond selected from the priority rule is different for CYS, {5β, 5α, 1β} In the main chain on the C-terminal side (dihedral angle position code BC), centering on the notation of one of the symbols {5β4β, 5β4α, 5α4β, 5α4α, 1β4β, 1β4α} In the case of a β sheet, the N-terminal main chain (dihedral angle position code A) is an exception because the sign {1α, 6β, 6α, 3β} (only the bond selected from the priority rule for CYS is different, { 2α, 5β, 5α}) and the main chain on the C-terminal side (dihedral angle position code BC In, {3β4β, 3β4α, 3α4β, 3α4α, 4β4β, 4β4α} is constructed around a one notation of the code of the secondary structure of the protein was found to be pattern recognition. Furthermore, detailed analysis has revealed that there are sites that can take very large secondary structure differences at amino acids 14-15 and 50-51. In comparison with the hypervariable region (the portion enclosed in the square in the amino acid sequence number of Table 3), one of the amino acid sequence sites that take a very large difference in secondary structure coincided with the CDR2 region. It was found that there is a site (amino acid sequence 14-15) that can take a very large difference in secondary structure even outside the region. The difference in structure between 2imm and 2mcp in amino acid sequence 14-15 is shown in FIG.

[Table 3] Encoded conformation notation and analysis results of immunoglobulin light chain (2imm, 2mcp)

(Example 3) Extraction of secondary structure patterns that allowed structural fluctuations in various immunoglobulin light chains and extraction of regions with extremely large changes other than CDRs
For the X-ray crystal structure data 2imm, 2mcp, 1a7o and 3iy2 of immunoglobulin light chain (MUS MUSCULUS, HOUSE MOUSE) registered in PDBj, using the method described in Patent Document 4, the molecular model accepts input. Based on the coded conformational notation consisting of a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. . A molecular model of the X-ray crystal structure data 1a7o is shown in FIG. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined.
Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. In order to deal with structural fluctuations, for one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining two surfaces The angle classification code was judged to have low homology, and was processed and compared for structural homology. The homology of the amino acid sequence was very low at 50.5%. In the sex comparison, the homology was 94.2%, the structural homology comparison of the main chain on the C-terminal side was 92.2% homology, and the entire main chain was 89.2% homology. It was confirmed that the homology of the secondary structure is almost the same as that of the case where the homology is very high. Table 4 shows the encoded conformation and homology analysis results of immunoglobulin light chain (2imm, 2mcp, 1a7o, 3iy2). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / permissible” and “main chain homology” refer to “swing” of the present invention. Represents the homology calculated by the “allowable calculation analysis method.” In addition, in the dihedral angle classification code in the table, the symbol “a” is used instead of the symbol “α” in order to reduce the processing load, and the symbol “β” is substituted. The symbol b is used.) The portion enclosed by a square indicates the amino acid sequence corresponding to the α helix and β sheet.
From this, it is possible to recognize the structure pattern of the secondary structure such as a random coil other than the α helix and β sheet by the encoded conformation notation, and the 52nd to 75th amino acids in the amino acid sequence of 2imm and 2mcp. SER THR ARG GLU SER GLY VAL PRO ASP ARG
"THR THR LEU ALA ASP GLY VAL PRO SER ARG PHE SER GLY SER GLY SER GLY THR GLN TYR SER LEU LYS ILE corresponding to PHE THR GLY SER GLY SER GLY THR ASP PHE THR LEU THR ILE" (SEQ ID NO: 3) ”and 3iy2“ LYS THR LEU THR GLU GLY VAL PRO SER ARG PHE SER GLY SER GLY SER GLY THR GLN PHE SER LEU LYS ILE (SEQ ID NO: 4) ” Able to extract amino acid sequence patterns that can take a protein structure with high structural homology in consideration of structural fluctuations even if it is low, including secondary structure patterns such as random coils other than α helix and β sheet all right.
Furthermore, detailed analysis shows that there are sites that can take very large secondary structure differences in amino acids 14-15, 30-31, 50-51, 76-77, and 92-94. It was issued. In comparison with the hypervariable region (the portion enclosed in the square in the amino acid sequence number of Table 4), three sites (30-31, 50-51, 92- 94th) corresponded to the regions of CDR1, CDR2 and CDR3, respectively. In addition, it was found that there are sites (14-15, 76-77) that can take very large secondary structure differences in framework regions that are not hypervariable regions. The difference in structure between 1a7o and 3iy2 in amino acid sequence 76-77 is shown in FIG.

[Table 4] Encoded conformational notation and analysis results of immunoglobulin light chain (2imm, 2mcp, 1a7o, 3iy2)

(Example 4) Comparison of homology allowing structural fluctuations in chaperonin GroEL having high amino acid sequence homology
For the X-ray crystal structure data 1kid and 1jon of chaperonin GroEL with high homology of amino acid sequences registered in PDBj, using the method described in Patent Document 4 above, each chemical bond Coding conversion was performed on the encoded conformation notation composed of a combination of a dihedral angle position code corresponding to a part and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. A molecular model of the X-ray crystal structure data 1kid is shown in FIG. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. Although the homology of the amino acid sequence is 100.0%, 1kid and 1jon are approximately 76.3% in the comparison of the structural homology of the N-terminal main chain according to the dihedral angle classification code determined by dividing 360 degrees into 12 parts. In comparison with the structural homology of the C-terminal main chain, the homology was only about 48.2%, and the protein structure was shaken in the X-ray crystal structure data.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining When the dihedral angle classification code was judged to have low homology, it was processed and compared for structural homology. The structural homology comparison of the main chain on the N-terminal side showed 99.3% homology and the main homology on the C-terminal side. The structural homology comparison of the strands also showed 99.3% homology and 97.9% homology of the entire main chain, indicating that homology comparison can be performed taking into account structural fluctuations. Table 5 shows the encoded conformation and homology analysis results of chaperonin GroEL (1kid, 1jon). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)
In addition, if the amino acid sequence corresponding to the α helix and β sheet enclosed in the square is processed in consideration of the structure fluctuation, each secondary structure can be expressed in the encoded conformation notation. For example, in the N-terminal main chain (dihedral angle position code A), the sign {6α, 3β, 3α} is an exception because only the bond selected from the priority rule is different for CYS, {5β, 5α, 1β} In the main chain on the C-terminal side (dihedral angle position code BC), centering on the notation of one of the symbols {5β4β, 5β4α, 5α4β, 5α4α, 1β4β, 1β4α} In the case of a β sheet, the N-terminal main chain (dihedral angle position code A) is an exception because the sign {1α, 6β, 6α, 3β} (only the bond selected from the priority rule for CYS is different, { 2α, 5β, 5α}) and the main chain on the C-terminal side (dihedral angle position code BC In, {3β4β, 3β4α, 3α4β, 3α4α, 4β4β, 4β4α} is constructed around a one notation of the code of the secondary structure of the protein was found to be pattern recognition.

[Table 5] Coordinate conformation notation and analysis result of chaperonin GroEL (1kid, 1jon)

(Example 5) In chaperonin GroEL with low homology of amino acid sequences, confirmation of secondary structural pattern allowing structural fluctuation and high structural homology
For the X-ray crystal structure data 1srv and 1kid of chaperonin GroEL registered in PDBj, the dihedral angles corresponding to each chemical binding site based on the molecular model that has been input using the method described in Patent Document 4 above. Coding conversion was performed on the coded conformational notation composed of a combination of a position code and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. A molecular model of the X-ray crystal structure data 1srv is shown in FIG. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. In order to deal with structural fluctuations, for one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining two surfaces The angle classification code was judged to have low homology and was processed and compared for structural homology. The structural homology of the main chain on the N-terminal side was compared even though the homology of the amino acid sequence was as low as 69.0%. , 99.3% homology, 99.3% homology in the structural homology comparison of the C-terminal main chain, and 97.9% homology in the entire main chain, the homology of the amino acid sequence of Example 4 It was confirmed that the homology of the secondary structure was almost the same as in the case of matching. Table 6 shows the encoded conformation and homology analysis results of chaperonin GroEL (1srv, 1kid). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / permissible” and “main chain homology” refer to “swing” of the present invention. Represents the homology calculated by the “allowable calculation analysis method.” In addition, in the dihedral angle classification code in the table, the symbol “a” is used instead of the symbol “α” in order to reduce the processing load, and the symbol “β” is substituted. The symbol b is used.) The portion enclosed by a square indicates the amino acid sequence corresponding to the α helix and β sheet.
From this, secondary structures such as random coils other than α-helix and β-sheet can be recognized by structural conformation, and the 193rd to 223th “TYR GLN” in the amino acid sequence of 1srv. PHE ASP LYS GLY TYR ILE SER PRO TYR PHE VAL THR ASN PRO GLU THR MET GLU ALA VAL LEU GLU ASP ALA PHE ILE LEU ILE VAL (SEQ ID NO: 5) “PRO TYR PHE ILE ASN LYS PRO GLU THR GLY ALA VAL GLU LEU GLU SER PRO PHE ILE LEU LEU ALA” (SEQ ID NO: 6) ” It was found that an amino acid sequence pattern capable of having a high protein structure can be extracted in a form including a secondary structure pattern such as a random coil other than an α helix and a β sheet.
Furthermore, it was extracted that there are some structural differences in amino acids 224 to 225 in the main chain structure. FIG. 13 shows the difference in structure between 1srv and 1kid in the amino acid sequence 224-225.

[Table 6] Coordinate conformation notation and analysis result of chaperonin GroEL (1srv, 1kid)

(Example 6) Application of “sway tolerance calculation analysis method” to X-ray crystal structure data of glutathione incorporated into glutathione-S-transferase and a molecular optimization molecular model by density functional theory
In the X-ray crystal structure data 6gst of glutathione-S-transferase (GST) registered in PDBj, two molecules of glutathione represented by the following chemical structural formula (I) are incorporated. Based on the molecular model that received the input using the method described in 4, the combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 A coding conversion was performed on the coded conformational notation. A molecular model of the X-ray crystal structure data 6gst_I is shown in FIG. At this time, in the dihedral angle position of the end of the cglu side chain and the end of the gly main chain, since the position of the hydrogen atom is not determined, the substituent to which the hydrogen atom is bonded cannot be determined. The preferred substituent for determining the corner classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
On the other hand, based on the conserved encoded conformational notation of 6gst_I, we predicted the missing hydrogen atom position, and created several glutathione (I) molecular models with different conformations, Structural optimization and frequency calculation were performed using the density functional method (B3LYP / 6-31G *). Among them, the most stable conformational structure was selected and defined to compare with X-ray crystal structure data.
(A) In accordance with the IUPAC nomenclature rules regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the bond plane is the smallest in the clockwise direction when viewed from the hydrogen atom.
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.
The coded conformation notation was stored in the database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. Despite the complete match of the amino acid sequences, 6gst_I and the molecular model obtained from the B3LYP calculation compare the structural homology of the main chain with the dihedral angle classification code determined by dividing 360 degrees into 12 However, the homology was not so high, and there was a fluctuation in the peptide thiol structure.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining The dihedral angle classification code was judged to have low homology, and was processed and compared for structural homology. Both structural homology comparison of the main chain on the N-terminal side and structural homology comparison of the main chain on the C-terminal side were performed. The homology was 100.0%, and it was found that the homology could be compared between the X-ray crystal structure data and the structure data obtained from the theoretical calculation in consideration of the structural fluctuation. A molecular model of glutathione (I) obtained from theoretical calculation is shown in FIG. Table 7 shows the encoded conformation notation and homology analysis results of glutathione (I) (6gst_I, B3LYP calculation). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)

[Table 7] Encoded conformational notation and analysis results of glutathione (I) (6gst_I, B3LYP calculation)

(Example 7) “Swing tolerance calculation analysis for molecular model obtained from X-ray crystal structure data and infrared circular dichroism (VCD) spectrum analysis of (S) -ibuprofen incorporated into human serum albumin (HSA)” Of law
In the X-ray crystal structure data 2bxg of human serum albumin (HSA) registered in PDBj, (S) -ibuprofen represented by the following chemical structural formula (II) of 4 molecules is incorporated, but 2bxg_A1, 2bxg_A2, 2bxg_B1 and 2bxg_B2 were determined by dividing the dihedral angle position code corresponding to each chemical binding site and 360 degrees into 12 parts based on the molecular model that accepted the input using the method described in Patent Document 4. Coding conversion was performed on the coded conformational notation composed of a combination of dihedral angle classification codes. In this case, in the isopropyl group of dihedral angle position code A, when (2) only one of the preferential substituents for determining the dihedral angle classification code is different from the other, it selects itself regardless of the order. Therefore, hydrogen was selected, but the encoding could not be performed because the position of the hydrogen atom was not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, in the carboxyl group of dihedral angle position code D, since the position of the hydrogen atom is not determined, the substituent to which the hydrogen atom is bonded cannot be determined, and the dihedral angle classification code for determining the dihedral angle classification code The preferred substituent could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in the database as character string text data.
On the other hand, in the deuterated chloroform solution obtained from the analysis using the density functional method (B3LYP / 6-31G *) of the infrared circular dichroism (VCD) spectrum described in Example 8 of Patent Document 4 above. The (S) -ibuprofen (II) conformational structure molecular model, which has a large abundance ratio, is newly defined to read from the stored database and compare it with the X-ray crystal structure data.
(A) In accordance with the IUPAC nomenclature rules regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the bond plane is the smallest in the clockwise direction when viewed from the hydrogen atom.
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.
Then, the coded conformation notation for X-ray crystal structure data comparison was stored in the database as character string text data.
Next, data was read by spreadsheet software, and processing was performed so that structural homology comparison could be performed for each dihedral angle position code. In the structural homology comparison for each dihedral angle position code by the dihedral angle classification code determined by dividing 360 degrees into 12, the X-ray crystal structure data and the structural data obtained from the X-ray crystal structure data and theoretical calculation There was not much homology with this combination, and there was shaking in the structure.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining The dihedral angle classification code was judged to have low homology and was processed and subjected to structural homology comparison. As a result, four X-ray crystal structure data (2bxg_A1,2bxg_A2,2bxg_B1,2bxg_B2) and theoretical calculation were obtained. Compared to the structural data (B3LYP calculation), the 12-fold homology is 25.0%, but when the calculation processing of the present invention is performed, the homology is evaluated as 75.0%, and the homology considering the fluctuation of the structure It was found that comparison was possible.
In addition, there are several conformational structures of (S) -ibuprofen (II) that exist multiple times in deuterochloroform solution, and the homologous structure considering the fluctuation of the structure is consistent. It was found that conformation notation can be extracted. Table 8 shows the results of the analysis of the homology of the (S) -ibuprofen (II) (2bxg_A1,2bxg_A2,2bxg_B1,2bxg_B2, B3LYP calculation) coding conformation and combinations thereof. (In the table, “homology / 12 division” represents the homology analyzed by the conventional method (Patent Document 4), and “homology” represents the homology calculated by the “swing allowable arithmetic analysis method” of the present invention. Also, in the dihedral angle classification codes in the table, in order to reduce the processing load, the code a is used instead of the code α, b is used instead of the code β, and code t is used instead of the code τ. The sign is used.)
Furthermore, in the structural data of the 2bxg_B1, 2bxg_B2 and B3LYP calculations, it was found that the homology considering the structural fluctuation was very high, but the conformation structure of the ppa-CD site was particularly different only in 2bxg_A1. .

[Table 8] Coordinate conformation notation of (S) -ibuprofen (II) (2bxg_A1,2bxg_A2,2bxg_B1,2bxg_B2, B3LYP calculation) and results of homology analysis by their combination

(Example 8) Structural homology comparison in consideration of fluctuation of structure in the presence or absence of interaction with immunoglobulin in prion protein
X-ray crystal structure data 1i4m of prion protein (HOMO SAPIENS, HUMAN) registered with PDBj and prion protein (HOMO SAPIENS, HUMAN) interact with immunoglobulin (MUS MUSCULUS, HOUSE MOUSE) For the data 2w9e, the dihedral angle position code corresponding to each chemical binding site and 360 degrees divided into 12 are determined based on the molecular model that has received the input using the method described in Patent Document 4. Coding conversion was performed on the coded conformational notation composed of combinations of face angle classification codes. The molecular models of the prion protein X-ray crystal structure data 1i4m and 2w9e are shown in FIGS. 16 and 17, respectively. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. Although the amino acid sequence was the same, the disulfide bond in the protein molecule was broken at 1i4m, and the structure of the whole molecule was very different as shown in FIGS. Furthermore, in the comparison of the structural homology of the N-terminal main chain according to the dihedral angle classification code determined by dividing 360 degrees into 12 for 1i4m and 2w9e, there is only about 48.0% homology, and the C-terminal main The structural homology comparison of the chains also showed a homology of only 42.9%, and in the X-ray crystal structure data, there was a fluctuation in the protein structure as well as the difference in the structure of the whole molecule.
Therefore, in order to eliminate the influence of structural fluctuation, for one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining The dihedral angle classification code was judged to have low homology, and was processed and compared for structural homology. The structural homology comparison of the main chain on the N-terminal side revealed a homology of about 91.8% and the C-terminal side. Comparison of the structural homology of the main chain of the protein was about 90.8% homology, and the whole main chain was about 86.9% of homology, indicating that structural change homology comparison can be performed in consideration of structural fluctuations. Table 9 shows the encoded conformation and homology analysis results of the prion protein (1i4m, 2w9e). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)
Therefore, with regard to amino acid sequences corresponding to other secondary structures such as α-helix, β-sheet, and random coil, each secondary structure is encoded conformational notation if arithmetic processing considering structural fluctuation is performed. It was found that the structure homology corresponding to the amino acid sequence was as high as 86.9% even though the structure was greatly different as a whole.
Furthermore, it was found from detailed analysis that there is a site capable of taking a very large secondary structure difference at amino acids 141-143 and 190-198. The difference in structure between 1i4m and 2w9e in amino acid sequence 141-143 is shown in FIG. In particular, as amino acid sequence patterns capable of changing the secondary structure of a large α helix, the 190 th to 193 th “THR THR THR THR (SEQ ID NO: 7)” and 1 i4m and 2w9e amino acid sequences in the 1i4m and 2w9e amino acid sequences Among them, the 195th to 198th “GLY GLU ASN PHE (SEQ ID NO: 8)” fragment was extracted. FIG. 19 shows the difference in structure between 1i4m and 2w9e in the amino acid sequence 190-198.

[Table 9] Encoded conformational notation and analysis results of prion protein (1i4m, 2w9e)

(Example 9) Comparison of structural homology in consideration of fluctuation of structure with and without interaction with prion protein in various immunoglobulins
X-ray crystal structure data 2mcp and prion protein (HOMO SAPIENS, HUMAN) of immunoglobulin (MUS MUSCULUS, HOUSE MOUSE) registered with PDBj interact with immunoglobulin (MUS MUSCULUS, HOUSE MOUSE) With respect to the structure data 2w9e, with respect to the entire region of the light chain including the variable region and the constant region in which the amino acid sequence is well conserved, each chemistry is determined based on the molecular model that has received input using the method described in Patent Document 4. Coding conversion was performed to the coded conformational notation composed of a combination of a dihedral angle position code corresponding to the binding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. FIG. 20 shows a molecular model of X-ray crystal structure data 2w9e in which prion protein and immunoglobulin interact. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. In 2mcp and 2w9e with 78.8% amino acid sequence homology, the structural homology comparison of the N-terminal main chain by dihedral angle classification code determined by dividing 360 degrees into 12 is only 52.1% homology Furthermore, even in the structural homology comparison of the main chain on the C-terminal side, there was only about 31.8% homology, and there was a fluctuation in the protein structure in the X-ray crystal structure data.
Therefore, in order to deal with the structural fluctuation, with respect to one dihedral angle classification code at the corresponding dihedral angle position, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining When the dihedral angle classification code was judged to have low homology, it was processed and the structural homology was compared. The structural homology comparison of the N-terminal main chain revealed that the homology of 89.6% In the structural homology comparison of the chains, the homology is 91.5% and the whole main chain is 84.4% homology. confirmed. Table 10 shows the encoded conformation and homology analysis results of the immunoglobulin light chain (2mcp, 2w9e). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 split” refers to homology analyzed by the conventional method (Patent Document 4), "Homology / shaking tolerance" and "main chain homology" represent the homology calculated by the "shaking tolerance calculation analysis method" of the present invention, and in the dihedral angle classification code in the table, (In order to reduce the load, the code a is used instead of the code α, and the code b is used instead of the code β.)
Detailed analysis reveals that in the variable region, amino acids 14-15, 29-31, 41-42, and 50-51 have sites that can take very large secondary structure differences. Two of them corresponded to the CDR1 and CDR2 regions of the hypervariable region (the portion surrounded by a square in the amino acid sequence number of Table 10). Furthermore, although the amino acid sequences of the constant regions are almost the same, the 150th to 158th “ILE ASP GLY SER GLU ARG GLN ASN GLY (SEQ ID NO: 9)” in the amino acid sequences of 2mcp and 2w9e It was found that there are sites that can take very large secondary structure differences between relatively long peptide chains. FIG. 21 shows the difference in structure between 2mcp and 2w9e in the amino acid sequence 150-158.

[Table 10] Encoded conformational notation and analysis results of immunoglobulin light chain (2mcp, 2w9e)

(Example 10) Application of “sway tolerance calculation analysis method” to an interaction site containing a side chain of a prion protein interacting with an immunoglobulin
Immunoglobulin (MUS MUSCULUS, HOUSE) whose amino acid sequence matches the X-ray crystal structure data 2w9e in which prion protein (HOMO SAPIENS, HUMAN) registered in PDBj interacts with immunoglobulin (MUS MUSCULUS, HOUSE MOUSE) MOUSE) only for X-ray crystal structure data 2w9d, extracting only the region of the light chain in which the prion protein and immunoglobulin interact, including the hypervariable region, and using the method described in Patent Document 4, The encoded conformation notation comprising a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 based on the molecular model that has received the input Coded conversion was performed. FIG. 22 shows a molecular model of the X-ray crystal structure data 2w9e in the region where the extracted prion protein interacts with the immunoglobulin light chain. At this time, when the dihedral angle position at the end of the valine (Val) side chain or leucine (Leu) side chain is the only preferred substituent for determining the dihedral angle classification code (2) In this case, hydrogen is selected because it is the case of selecting itself regardless of the order, but the encoding cannot be performed because the position of the hydrogen atom is not determined. Therefore, when the X-ray crystal structure data were compared, (B) the hydrogen atom that should exist originally was placed in front of the paper, and the bond of the dihedral position code was placed perpendicular to the paper on the other side of the paper. A substituent or an atom whose dihedral angle due to a surface formed by a bond of a hydrogen atom and a dihedral angle position code or a surface formed by a bond of an atom and a dihedral angle position code is the smallest clockwise as viewed from the hydrogen atom Coded conversion was performed according to the priority rule of selecting. In addition, hydrogen atoms are bonded because the position of the hydrogen atom is not determined at the dihedral angle positions of the arginine (Arg) side chain, aspartic acid (Asn) side chain, and glutamic acid (Glu) side chain. One of the substituents could not be determined, and the preferred substituent for determining the dihedral angle classification code could not be selected. Therefore, when the substituent or atom preferred when (C) other hydrogen atoms are excluded cannot be determined as determined when comparing the X-ray crystal structure data, the substitution with the smallest dihedral angle is performed. Coded conversion was performed according to the priority rule of selecting groups or atoms. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Next, the data is read by spreadsheet software, and for each amino acid sequence, the structural homology can be compared by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. Processed. In 2w9e and 2w9d, which have the same amino acid sequence, the structural homology of the main chain on the N-terminal side according to the dihedral angle classification code determined by dividing 360 degrees into 12 is only about 72.2% homology, and the C-terminal The structural homology comparison of the main chain on the side showed only about 51.9% homology, and there was fluctuation in the protein structure in the X-ray crystal structure data.
Therefore, in order to cope with structural fluctuation and to compare the structural homology of the side chains, one dihedral angle classification code at the corresponding dihedral angle position is as follows: 1) Dihedral angle adjacent to the dihedral angle classification code Combination of classification codes has high homology (○), 2) Moderate homology of relatively close dihedral angle classification codes across one dihedral angle classification code (△), 3) Other When the homology of the dihedral angle classification code combination was judged to be low (×) and the structural homology comparison was performed, the structural homology comparison of the N-terminal main chain showed 98.1% homology, Compared to the structural homology of the main chain on the C-terminal side, the homology is 100.0%, and even if the prion protein interacts with the homology comparison considering structural fluctuations, it is near the hypervariable region of the immunoglobulin light chain. No major secondary structure changes in the main chain It has been confirmed. Table 11 shows the encoded conformation and homology analysis results of the extracted regions of the immunoglobulin light chain (2w9e, 2w9d). (In the table, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / (strict)” refers to homology analyzed by the conventional method (Patent Document 4). , “Homology / shaking tolerance” represents the homology calculated by the “shaking tolerance calculation analysis method” of the present invention, and α in order to reduce the processing load in the dihedral angle classification code in the table. (The code a is used instead of the code, and the code b is used instead of the code β.)
From the detailed analysis, it was found that even when the amino acid side chain of the immunoglobulin light chain was taken into account in the structural fluctuation, even when the prion protein interacted, it had 78.7% structural homology. In particular, in the 31st TYR, 49th ASP, 90th TRP, 91th ARG and 95th TYR that interact directly with the prion protein, two dihedral angle classification codes of the 91st ARG Except for, it was found that the structural homology was high considering the fluctuation of the structure. Furthermore, the site | part which can take a big structural difference in an amino acid side chain was extracted. The difference in structure between 2w9e and 2w9d in amino acid sequence 90-92 is shown in FIG.
Furthermore, on the prion protein side of the extracted region, the structural homology comparison of the amino acid side chains of 2w9e and 1i4m was also performed using the analysis results of Example 8. Table 12A (main chain) and Table 12B (side chain) show the encoded conformational notation and homology analysis results of the extracted region of the prion protein (2w9e, 1i4m). (In Table 12A, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / (strict)” refers to homology analyzed by the conventional method (Patent Document 4). The sex, “homology / accuracy” represents the homology calculated by the “sway tolerance calculation analysis method” of the present invention.In Table 12B, the first “homology / aa sequence” is “amino acid sequence homology ( Identity ”),“ homology / (strict) ”is the homology analyzed by the conventional method (Patent Document 4), and“ homology / shaking ”is the“ shaking tolerance calculation analysis method ”of the present invention. In addition, in the dihedral angle classification codes in Tables 12A and 12B, the sign a is substituted for the sign α and b is substituted for the sign β in the dihedral angle classification codes in Tables 12A and 12B. In particular, the number 148 interacts directly with the immunoglobulin light chain. In the ARG and 151 th ARG, structural homology also not without side chain only the main chain in view of the sway of the structure was found to be very high. In addition, in the comparison of dihedral angle classification codes of only the base part of the amino acid side chain, a structural homology of 92.0% was found in the extracted region in consideration of the structural fluctuation. The difference in structure between 2w9e and 1i4m in the amino acid sequence 151-152 is shown in FIG.

[Table 11] Encoded conformational notation and analysis result of immunoglobulin light chain (2w9e, 2w9d) of extracted region

[Table 12A] Encoded conformational notation and analysis results (principal chain) of prion protein (2w9e, 1i4m) in the extracted region
[Table 12B] Encoded conformational notation and analysis result (side chain) of prion protein (2w9e, 1i4m) in the extracted region

(Example 11) Analysis of α-helix and β-sheet structure pattern in free factor RF2 X-ray crystal structure data 1i4m and prion protein (HOMO SAPIENS) of prion protein (HOMO SAPIENS, HUMAN) rich in α-helix fragment of Example 8 , HUMAN), 3β5α4β was extracted as the α-helix structure pattern of the main chain (ABC).
On the other hand, the X-ray crystal structure data 2mcp of immunoglobulin (MUS MUSCULUS, HOUSE MOUSE) and prion protein (HOMO SAPIENS, HUMAN), which are rich in β-sheet fragments of Example 9, interact with immunoglobulin (MUS MUSCULUS, HOUSE MOUSE). From the acting X-ray crystal structure data 2w9e, 6α4β4β was extracted as the β-sheet structure pattern of the main chain (ABC).
Next, the ribosomal RNA (THERMUS THERMOPHILUS) registered in PDBj interacts with the release factor RF2 (THERMUS THERMOPHILUS). X-ray crystal structure data 2wh1 and the release factor RF2 (THERMUS THERMOPHILUS) have high amino acid sequence homology. Only X-ray crystal structure data 2ihr was subjected to coding conversion using the same method as in Example 8. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Data is read by spreadsheet software and processed so that the structure homology can be compared for each amino acid sequence by separating the main chain on the N-terminal side, the main chain on the C-terminal side, and the side chain across the side chain. It was. 2ihr and the α-helix structure pattern of the main chain (ABC) 3-7. As shown in Table 13, it was confirmed that there was about 59.4% α-helix structure pattern as shown in Table 13, and in other areas, other than α-helix. The structure pattern was extracted. Similarly, for 2ihr, the above 3-7. With the β sheet structure pattern of the main chain (ABC) As shown in Table 14, it was confirmed that there was about 30.0% of the β sheet structure pattern, and in other areas, the structure other than the β sheet was analyzed. A pattern was extracted.
Furthermore, in order to extract only the difference in the large conformational structure, the structure data 2ihr when the free factor RF2 alone is used and the structure data of 2wh1 where a relatively large fluctuation in the molecular structure is expected due to the interaction with the ribosomal RNA molecule 3-7. The structural homology was evaluated by applying the improved "sway tolerance calculation method". According to the structural homology evaluation, as shown in Table 15, it was confirmed that the structural homology was about 88.9%. (In Tables 13 to 15, the first “homology / aa sequence” refers to “amino acid sequence homology (identity)”, and “homology / 12 division” is determined by the conventional method (Patent Document 4). The analyzed homology, “main-chain homology”, represents the homology calculated by the “swing tolerance calculation analysis method” improved in 3-7 of the present invention. In the angle classification code, in order to reduce the processing load, the code a is used instead of the code α, and the code b is used instead of the code β.) In addition, the α helix previously extracted is used. The structural change fragments of the main chain in the structural pattern other than 2ihr α-helix and β-sheet were extracted in combination with the structural pattern other than 1 and structural pattern analysis other than β-sheet (Table 15). FIG. 25 shows the difference in the structure of the whole molecule of the free factor RF2 in 2ihr and 2wh1 due to the superposition of the molecular models. As shown in FIG. 25, in 2ihr and 2wh1, structural changes occur at a plurality of locations at the same time, and it is difficult to grasp the difference in structure by analysis by superimposing conventional molecular models. From the analysis results shown, it was proved that the structural change region of the fragment that essentially functions can be understood by coding in the molecular structural change caused by the interaction with other molecules.

[Table 13] Coding conformation notation of release factor RF2 (2ihr) and α helix structure pattern analysis result

[Table 14] Coding conformation notation of release factor RF2 (2ihr) and β sheet structure pattern analysis result

[Table 15] Coding conformation notation and analysis results of free factor RF2 (2ihr, 2wh1)

(Example 12) Confirmation of stem-type structural pattern in ribosomal RNA and extraction of structural pattern characteristic of structural change of ribosomal RNA caused by difference between elongation factor EF-Tu and release factor RF2
Definition of dihedral angle position codes of fragments constituting RNA represented by the following chemical structural formula (III) for Xr-ray crystal structure data 1r9f of synthesized RNA having stem type double helix structure registered in PDBj Based on the above, coding conversion was performed using the same method as in Example 8. From 1r9f, 3β1β3β-1α-1α3β was extracted as the stem-type structure pattern of the main chain (ABC-D-EF).
Next, the ribosome registered in PDBj (THERMUS THERMOPHILUS) interacts with the elongation factor EF-Tu (THERMUS THERMOPHILUS). Regarding the molecular structure data of each ribosomal RNA of the interacting X-ray crystal structure data 2wh1, based on the definition of the dihedral angle position code of the fragment constituting the RNA represented by the following chemical structural formula (III), Example 8 The coding conversion was performed using the same method. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Data was read by spreadsheet software, and processing was performed so that the structure homology comparison could be performed by separating the main chain and side chain for each nucleobase sequence. 2wrn with the main chain (ABC-D-EF) stem-type structure pattern 3β1β3β-1α-1α3β, 3-7. As a result of analysis using the improved “swing tolerance calculation analysis method”, about 63.6% of the stem type structural patterns existed, and structural patterns other than the stem type were extracted.
Further, in order to extract only the difference in the large conformational structure, the structure data of ribosomal RNA in 2wrn and 2wh1 are described in 3-7. The structural homology was evaluated using the improved "sway tolerance calculation analysis method". 3-7. As shown in Table 16, it was confirmed that structural homology of about 88.5% was obtained in the structural homology evaluation using the improved “swing tolerance calculation analysis method”. (In the table, the first “homology / base sequence” refers to “nucleobase sequence homology (identity)”, and “homology / 12 division” refers to homology analyzed by the conventional method (Patent Document 4). , “Homology of main chain” represents the homology calculated by the improved “swing tolerance calculation analysis method” of 3-7 of the present invention. In order to reduce the load, the code a is used in place of the code α and the code b is used in place of the code β.) In combination, structurally altered fragments of the main chain in a structural pattern other than the 2wrn stem type were extracted (Table 16). FIG. 26 shows the difference in the structure of the entire molecule of ribosomal RNA in 2wrn and 2wh1 due to the superposition of molecular models. As shown in FIG. 26, when a giant biopolymer such as ribosomal RNA is used, it is difficult to grasp the difference in structure by analysis by conventional molecular model superposition, but the analysis results shown in Table 16 From the above, it was proved that the structural change region of the fragment that essentially functions in the molecular structural change caused by the interaction with other molecules can be understood by coding.
Furthermore, when a detailed analysis was performed focusing on the peripheral part interacting with the elongation factor EF-Tu shown in FIG. 27 at 2wrn, A55, C366-U368 and C372 each had an A -1β5β2α-6α-3β6α, C-3α1β3β-6α-3β1α-U-2α1α2β-6α-3β3β-U-3β5β2α-6α-6β2α and C-3β5β2β-6α-6α1β main chain structural pattern, only 2wh1 It was found to be an RNA fragment that can take a structural pattern different from the structure obtained. FIG. 28 shows the difference in structure at C366-C372 of ribosomal RNA in 2wrn and 2wh1 by superimposing molecular models. In this way, it is difficult to grasp the difference in structure by the conventional analysis by superimposing molecular models, but there are characteristic structural patterns related to the function in the molecular structure change caused by the interaction with other molecules. It was proved that it can be understood by coding.

[Table 16] Ribosomal RNA
(2wrn, 2wh1) coding conformation notation and analysis results

(Example 13) Extraction of characteristic new structure pattern in small interfering RNA (siRNA)
Small interfering RNA (siRNA) X-ray crystal structure data registered in PDBj 1di2_C, 1di2_D, 1di2_E, 1di2_G, 1r9f_B, 1r9f_C, 1rpu_CA, 1rpu_CB, 1rpu_DA, 1rpu_DB, 1si2_B, 1si3_B, 1b 2bgg_R, 2bgg_S, 2f8s_C, 2f8s_D, 2f8t_C, 2f8t_D, 2g92_A, 2g92_B, 3cz3_EA, 3cz3_EB, 3cz3_FA, 3cz3_FB, 3cz3_GA, 3cz3_GB, 3cz3_H, 3cz3 The coding conversion was performed based on the definition of the face angle position code. The obtained encoded conformation notation for extracting a new structure pattern was stored in a database as character string text data.
Stem structure of main chain (ABC-D-EF) extracted from the encoded conformation notation for extracting a new structure pattern and 1r9f of Example 12 stored in a database by spreadsheet software The pattern (3β1β3β-1α-1α3β) data was read and processed so that the structure homology could be compared by separating the main chain and side chain for each nucleobase sequence. Regarding the encoded conformational notation of siRNA for extracting a new structural pattern, the 3-7 above with the main chain (ABC-D-EF) stem type structural pattern 3β1β3β-1α-1α3β. As a result of analysis using the improved `` swing tolerance calculation analysis method '', there are about 72.6% of structural patterns that are highly homologous to the stem type structural pattern 3β1β3β-1α-1α3β, and other structural patterns are extracted. It was.
From the other structural patterns, a structural pattern 1β1α2β-1α-6α3β having a high appearance frequency is extracted, and the above 3-7. Using the improved “sway tolerance calculation analysis method”, it was confirmed that the structure pattern is different from the stem-type structure pattern 3β1β3β-1α-1α3β. Next, with respect to the coded conformational notation of siRNA for extracting a new structure pattern, the 3-7 above with the structure pattern 1β1α2β-1α-6α3β. As a result of analysis using the improved “swing tolerance calculation method”, about 10.8% of structural patterns having high homology with the structural pattern 1β1α2β-1α-6α3β existed, and other structural patterns were extracted.
Among the structural patterns other than the stem-type structural pattern 3β1β3β-1α-1α3β and the structural pattern 1β1α2β-1α-6α3β, 3-7. After confirming that the pattern was a new structure pattern using the improved “swing tolerance calculation analysis method” of the above, when the new structure pattern extraction process was repeated, 21 structure patterns were extracted and analyzed. All X-ray crystal structure data of siRNA had a structure pattern belonging to any of these 21 structure patterns. Table 17 shows the classification of the 21 extracted structural patterns in descending order of appearance frequency. (In the dihedral angle classification codes in the table, the symbol a is used instead of the symbol α, and the symbol b is used instead of the symbol β.)

[Table 17] Classification of structural patterns found in siRNA

(Example 14) Extraction and evaluation of characteristic new structure pattern in RNA having SARCIN-RICIN LOOP
For X-ray crystal structure data 2x7n_A of RNA having SARCIN-RICIN LOOP registered in PDBj (SACCHAROMYCES CEREVISIAE, BAKER'S YEAST, FRAGMENT 2684-2711), The coding conversion was performed based on the definition of dihedral angle position code. The encoded conformational notation for comparing the obtained X-ray crystal structure data was stored in a database as character string text data.
Using spreadsheet software, read the encoded conformational notation stored in the database and the 21 structural patterns extracted in Example 13, and separate the main and side chains for each nucleic acid base sequence. Processing was performed so that structural homology comparison could be performed. Regarding the above encoded conformational notation of 2x7n_A, the above 3-7 with the main chain (ABC-D-EF) stem type structural pattern 3β1β3β-1α-1α3β. As a result of analysis using the improved `` swing tolerance calculation analysis method '', there are about 74.1% of structural patterns highly homologous to the stem type structural pattern 3β1β3β-1α-1α3β, and other structural patterns are extracted It was.
For the other structural patterns, the classification of the remaining 20 structural patterns found in siRNA is used, and the above 3-7. As a result of analysis using the improved “swing tolerance calculation analysis method”, one structural pattern having high homology with structural pattern 5 shown in Table 17 of Example 13 and high homology with structural pattern 6 were obtained. There were two structural patterns and one structural pattern having high homology with structural pattern 10. In A8 and G9, the main chain structural patterns A-1β5α3β-3β-6α1β and G-3β2α2β-6α-1α5α characteristic of SARCIN-RICIN LOOP were found. As shown in Table 18, these two new structural patterns include 21 structural patterns found in siRNA and 3-7. The structural homology was not confirmed at all even when the homology was evaluated using the improved “sway tolerance calculation analysis method”. Table 19 shows the structure pattern analysis results of 2x7n_A. (In Table 18 and Table 19, the first “homology / 12 division” is the homology analyzed by the conventional method (Patent Document 4), and the “homology of the main chain” is the above-mentioned 3-7. This represents the homology calculated by the improved “swing tolerance calculation analysis method.” In addition, in the dihedral angle classification codes in Table 18 and Table 19, the code “a” is used instead of the code “α” in order to reduce the processing load. The symbol b is used in place of the symbol β.) Based on the analysis results shown in Table 19, the structural pattern of the main chain characteristic of SARCIN-RICIN LOOP is encoded without using any graphical interface. It was proved that it can be understood. FIG. 29 shows a 2 × 7n_A molecular model for showing the structural pattern of the main chain characteristic of SARCIN-RICIN LOOP. (In the molecular model in the figure, the whole molecule is shown using a ribbon display, and only the new structure pattern of the main chain characteristic of SARCIN-RICIN LOOP is shown using a stick display.)

[Table 18] Evaluation of structural homology of the main chain between the structural pattern seen in siRNA and the structural pattern characteristic of SARCIN-RICIN LOOP

[Table 19] Encoded conformational notation and structural pattern analysis results of RNA (2x7n_A) with SARCIN-RICIN LOOP

  The conformational homology evaluation apparatus according to the present invention is capable of dealing with one of the two dihedral angle positions at a corresponding dihedral angle position even when a simple arithmetic process based on the coincidence of codes cannot completely cope with the problem of fluctuation of the molecular structure. Regarding the face angle classification code, the combination of the dihedral angle classification code and the adjacent dihedral angle classification code is highly homologous, and the remaining dihedral angle classification codes are judged to be low in homology, and are processed. Therefore, the conformational structure homology can be easily evaluated in consideration of the problem of fluctuation of the molecular structure based on the equilibrium. Therefore, for example, there is a lyophilization step in the liquid preparation process of the antibody drug, but the structure using the encoded conformation notation indicates that the structure change of the antibody drug has not occurred in the solution before and after the lyophilization. It can be easily confirmed by homology comparison.

Further, according to the present invention, the molecular model obtained from the X-ray crystal structure analysis without the position information of the hydrogen atom can be compared with the molecular model obtained from the theoretical calculation, and the analysis target compound is a protein. Alternatively, an amino acid sequence pattern having a very high structural homology can be extracted together with a detailed conformation notation in the case of a peptide, although the amino acid sequence is different. In addition, regarding the interaction site between the ligand molecule and the protein molecule, the structural change can be simply described by detailed conformational notation including amino acid side chains by structural homology evaluation. Furthermore, analysis of characteristic structural patterns of fragments such as α-helix type and β-sheet type can be performed with a simple operation using a code. Similarly, in various RNAs, a structure pattern corresponding to a characteristic secondary structure such as a stem structure or a loop structure can be extracted, and a new structure pattern in a protein, RNA or the like can be found.
Therefore, for example, the structure homology analysis of an orphan receptor protein whose function has not yet been clarified can be used for elucidation of the function, so that it can be used for designing a new drug.

  Therefore, the present invention is sufficiently developed in applications such as quality control methods that guarantee further safety of biopharmaceuticals, drug discovery by pharmacological proteomics, and assessment of the effects of harmful substances on the human body. Is expected.

SEQ ID NO: 1: Insulin A chain SEQ ID NO: 2: Immunoglobulin light chain (52-75aa) 2imm, 2mcp
Sequence number 3: Immunoglobulin light chain (52-75aa) 1a7o
Sequence number 4: Immunoglobulin light chain (52-75aa) 3iy2
Sequence number 5: Chaperonin GroEL (193-223aa) 1srv
Sequence number 6: Chaperonin GroEL (193-223aa) 1kid
Sequence number 7: Prion protein (190-193aa) 1i4m, 2w9e
Sequence number 8: Prion protein (195-198aa) 1i4m, 2w9e
Sequence number 9: immunoglobulin light chain (150-158aa) 2mcp, 2w9e

Claims (10)

The processing unit accepts input of a molecular model of the conformation structure of each of a plurality of analysis target compounds,
Based on the molecular model for which the processing unit has received the input, the processing unit includes a three-dimensional configuration composed of a combination of a dihedral angle position code corresponding to each chemical bonding site and a dihedral angle classification code determined by dividing 360 degrees into 12 parts. Note that the conformation notation is necessary for a structure in which the conformation can be uniquely determined by one conformation notation by means of the conformation notation (hereinafter referred to as an encoded conformation notation). Extracting the encoded conformation notation that leaves the notation of the chemical bonding site and omits unnecessary notation from the molecular model, and stores the extracted encoded conformation notation in the storage unit,
The processing unit retrieves a plurality of the encoded conformational representations from the storage unit;
At each position in the corresponding dihedral angle position, the dihedral angle classification code on both sides is extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrast elements of the set at each position in each analyte compound;
The condition that at least one element among the elements of the set belonging to each compound is common to all compounds, and that the common element is the central element in the set belonging to at least one compound. Search for whether or not
A conformation homology evaluation method characterized by repeating an operation of performing a calculation process for determining that the homology is high when satisfying and low otherwise. According to the standard rule for determining the dihedral angle classification code based on the IUPAC nomenclature, the coordinate position is at least one conformational structure of the compound to be compared is a molecular model by X-ray crystal structure analysis. The coded conformational notation is selected according to the following priority rules (A) to (C) when a dihedral angle classification code involving a hydrogen atom cannot be determined due to being undefined: Conformational homology evaluation method according to 1,
(A) In accordance with the rules of IUPAC nomenclature regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the surface formed by the bond is the smallest clockwise when viewed from the hydrogen atom,
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.   When the target compound to be compared is a protein or peptide, the encoded conformation corresponding to the amino acid sequence is selected based on the main chain that binds each amino acid residue and using each amino acid residue as a prefix. The conformation homology evaluation method according to claim 1 or 2, wherein the structure homology is evaluated only for the main chain.   The three-dimensional conformation according to claim 3, wherein the encoded conformation notation and amino acid sequence pattern having high structural homology but different amino acid sequences are extracted between proteins or peptides to be analyzed. Conformational homology evaluation method.   In the case where the analysis target compounds to be compared have the same fragment, the notation of the dihedral angle position code including the prefix of the fragment is completely the same, and one of them can be selected as a reference compound In this case, the dihedral angle classification code of the conformation structure of the compound is the central code among the three consecutively adjacent ones, and the dihedral angle at the corresponding dihedral angle position of the compound for which structural homology evaluation is performed If all the elements of the set of angle classification codes match the median value of the standard dihedral angle classification code or the dihedral angle classification code adjacent to it, the homology is high. The conformation homology evaluation method according to claim 1, wherein the calculation processing is performed after determination. A conformation homology evaluation apparatus comprising a display unit, an input unit, a processing unit, a main storage unit, an output unit, an I / O unit, a database unit, and a bus connecting each unit,
In the input unit, in accordance with the input operation, the molecular model of the conformation structure of each of the plurality of analysis target compounds is extracted from the database unit and sent to the processing unit,
In the processing unit, for each molecular model that has received an input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, each of the encoded conformational representations corresponding to the plurality of analysis target compounds is extracted from the storage unit,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrast elements of the set at each position in each analyte compound;
The condition that at least one element among the elements of the set belonging to each compound is common to all compounds, and that the common element is the central element in the set belonging to at least one compound. Search for whether or not
If it satisfies, it is determined that the homology is high, otherwise it is determined that the homology is low,
Subsequently, the same operation at the next position is performed, and when the operation for all corresponding positions is repeated and sent to the output unit,
The conformation homology evaluation apparatus , wherein the output unit outputs a processing result generated by the processing unit. In determining the encoded conformational notation for each molecular model in the processing unit, at least one of the molecular models that has received the input is a molecular model by X-ray crystal structure analysis, and IUPAC nomenclature When the dihedral angle classification code involving hydrogen atoms cannot be determined because the coordinate position is not determined according to the standard rule for determining the dihedral angle classification code in accordance with the above, the following priority rules (A) to (C) are followed. A conformation homology evaluation apparatus according to claim 6, further comprising a step of selecting the encoded conformation notation;
(A) In accordance with the rules of IUPAC nomenclature regarding conformation, when a preferred substituent or atom can be determined even when a hydrogen atom is removed, the preferred substituent or atom is selected.
(B) Conformational IUPAC nomenclature (2) When only one is different from the other, when hydrogen is selected according to the rule of selecting itself regardless of the order, Place the bond of the dihedral angle position code in front of the page, and place the bond of the dihedral angle code perpendicular to the surface of the sheet on the other side of the sheet. Select the substituent or atom whose dihedral angle due to the surface formed by the bond is the smallest clockwise when viewed from the hydrogen atom,
(C) If the preferred substituent or atom cannot be determined when other hydrogen atoms are removed, the substituent or atom having the smallest dihedral angle is selected.   In the case where the analysis target compounds to be compared have the same fragment, the notation of the dihedral angle position code including the prefix of the fragment is completely the same, and one of them can be selected as a reference compound In this case, the dihedral angle classification code of the conformation structure of the compound is the central code among the three consecutively adjacent ones, and the dihedral angle at the corresponding dihedral angle position of the compound for which structural homology evaluation is performed If all the elements of the set of angle classification codes match the median value of the standard dihedral angle classification code or the dihedral angle classification code adjacent to it, the homology is high. The conformation homology evaluation apparatus according to claim 6 or 7, characterized in that arithmetic processing for determination is performed. The processing unit accepts an input of a molecular model of a conformation structure of a fragment corresponding to one or more reference structural patterns,
The processing unit describes the dihedral angle position code corresponding to the chemical bonding site of interest for the molecular model that has received the input, and 360 degrees is divided into 12 for each dihedral angle position code. By applying the encoded conformation notation that combines the specified dihedral angle classification codes, the conformation regarding each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
The extracted encoded conformation notation is stored in advance in the storage unit as a reference structure pattern,
Separately, accepting the input of the molecular model of the conformational structure of the compound to be analyzed,
In the processing unit, for the molecular model of the analysis target compound that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Next, an extracted conformational notation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for the structural pattern analysis is extracted,
Each extracted conformation conformation notation is stored in the storage unit,
In the processing unit, the encoded conformational notation of the analysis target compound is taken out from the storage unit, and the reference structure pattern to be compared is selected from the reference structure patterns stored in the storage unit in advance. , Taking out the encoded conformational notation that matches the structure pattern as a reference for comparison,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrasting elements of the set at the positions in the structural pattern to be analyzed and the reference structural pattern,
Search for whether or not the condition that at least one element is common to the reference structure pattern among the elements of the set to which each dihedral angle classification code belongs,
If it satisfies, it is determined that it matches the reference structure pattern, and if it does not, it is determined that it corresponds to the other structure pattern.
Subsequently, an operation for performing the same calculation process at the next position is performed, and the operation for all corresponding positions is repeated. A conformation homology evaluation apparatus comprising a display unit, an input unit, a processing unit, a main storage unit, an output unit, an I / O unit, a database unit, and a bus connecting each unit,
In the input unit, in accordance with an input operation, accepts an input of a molecular model of a conformation structure of a fragment corresponding to one or more reference structure patterns,
The molecular model of the conformational structure of each structural pattern is taken out from the database part and sent to the processing part,
In the processing unit, for each molecular model of each structural pattern that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
The extracted encoded conformation notation is stored in advance in the storage unit as a reference structure pattern,
Separately, accepting the input of the molecular model of the conformational structure of the compound to be analyzed,
In the processing unit, for the molecular model of the analysis target compound that received the input,
Encoding that expresses a dihedral angle position code corresponding to the chemical bonding site of interest and combines each dihedral angle position code with a dihedral angle classification code determined by dividing 360 degrees into 12 parts. By giving conformation notation, the conformation about each molecular model is uniquely determined,
Then, extract the conformation conformation that omits unnecessary notation other than the notation of the chemical bonding site of interest necessary for the calculation for structural homology evaluation,
Each extracted conformation conformation notation is stored in the storage unit,
The processing unit extracts the encoded conformational notation of the analysis target compound from the storage unit, and selects a reference structure pattern to be compared from the reference structure patterns stored in the storage unit in advance. And taking out the encoded conformational notation that matches the structure pattern as a reference for comparison,
At the first position in the corresponding dihedral angle position, the dihedral angle classification codes on both sides are extracted together with the selected dihedral angle classification code, and the three dihedral angles that are consecutively adjacent as elements are extracted. Create a set with classification code as an element,
Contrasting elements of the set at the positions in the structural pattern to be analyzed and the reference structural pattern,
Search for whether or not the condition that at least one element is common to the reference structure pattern among the elements of the set to which each dihedral angle classification code belongs,
If it satisfies, it is determined that it matches the reference structure pattern, and if it does not, it is determined that it corresponds to the other structure pattern.
Subsequently, an operation for performing the same calculation process at the next position is performed, and when the operation for all corresponding positions is completed repeatedly, it is sent to the output unit,
The conformation homology evaluation apparatus , wherein the output unit outputs a processing result generated by the processing unit.