JPH0740014B2 - DNA pattern reader - Google Patents

Description

TECHNICAL FIELD The present invention relates to a DNA pattern reader, and more particularly, to an apparatus for reading a DNA sequence film image and automatically determining the base sequence of a gene. The present invention relates to a DNA pattern reader suitable for discriminating boundaries of bands (lanes) corresponding to four bases of C and T.

[Conventional technology]

In a device that recognizes the base sequence of a gene from the image of a film (DNA sequence film) that has been labeled with a radioisotope and subjected to electrophoresis, A, G, C, T developed in the direction of electrophoresis The identification of the boundaries of bands (lanes) corresponding to one base is the most basic factor that determines the recognition accuracy of the base sequence of a gene. A conventional device of this type, for example, as in Japanese Patent Application No. 60-241548 (Japanese Patent Laid-Open No. 62-102375), simply adds image data of a DNA sequence film in the direction of electrophoresis, The boundaries of bands (lanes) corresponding to four bases were identified by detecting this kind of change in the direction perpendicular to the direction of the electrophoresis.

[Problems to be solved by the invention]

On the DNA sequence film, the images (bands) of individual bases form elongated colonies perpendicular to the electrophoretic direction, and the slots for each base appear in a staggered manner. In the conventional technology, no consideration is given to this point, and image data is simply added in the migration direction.
There is a problem that it is difficult to detect boundaries between base lanes in which the amount of band colonies is approximately equal.

An object of the present invention is to positively utilize the information on the deviation of band colonies, to clearly detect the fault between base lanes even when the amount of band colonies is close, and to realize higher recognition accuracy. .

[Means for solving problems]

The above-mentioned object is, in the step of acquiring and processing the image of the DNA sequence film, with respect to the image data in the direction (X-axis direction) perpendicular to the electrophoretic direction (Y-axis direction), the variation (difference between adjacent pixels) ), The absolute value is added to each coordinate on the X-axis in the Y-axis direction, and the X-coordinate position at which the addition result (sum of difference absolute values) becomes maximum is A, G, C, T.
It is achieved by recognizing the boundary of the band (lane) corresponding to each base of.

[Work]

The circuit for obtaining the sum of absolute differences is operated in parallel with the operation of sending digitized image data to a two-dimensional memory such as a computer. As a result, no extra processing time is required as compared with the case where software is applied to the image data on the memory to determine the boundaries between the base lanes, and the efficiency of the image reading apparatus is not reduced. Also, DN
Although there are individual variations in the image density on the A-sequence film, the circuit built into the device efficiently sums many points, so the statistical result is a distribution with peaks at the lane boundaries, Since the noise portion is removed, the rate of erroneous recognition is low, and even when a peculiar peak with overlapping noise appears, software recovery can be performed by looking at the interval, so no malfunction occurs.

〔Example〕

 An embodiment of the present invention will be described below with reference to the drawings.

First, the principle of boundary recognition of the lane corresponding to each base according to the present invention will be described with reference to FIG. In FIG. 2, 10 is a film (DNA sequence film) as a result of electrophoresis after labeling with a radioactive isotope or the like, and 101 to 104 are A, G,
The bands (lanes) corresponding to the four bases of C and T, and 105 are the base image images (bands) due to radioactive isotopes.

The scanning is performed with the direction (X-axis direction) perpendicular to the electrophoresis direction of the DNA sequence film 10 as the main scanning, and sequentially moves in the sub-scanning direction, that is, the electrophoresis direction (Y-axis direction).
Here, the band 105 on the film 10 is an image obtained by electrophoresing nucleotides having different molecular weights, and each has an elongated shape in the X-axis direction, and the image density is almost uniform in the central part of the band and at both ends. It changes a lot. Then, when the absolute value (difference absolute value) of the difference in image density (image data) between adjacent pixels in the main scanning direction (X-axis direction) obtained by scanning is taken, the value is the center of the band 105. It is almost zero in the part and large near both ends. A value obtained by adding the absolute difference value in the Y-axis direction for each same X-axis coordinate for each main scanning line will be referred to as a sum of absolute difference values. This sum of absolute differences is X of the base pixel (band 105) on the film 10.
A characteristic quantity representing a change in the axial direction is shown, and as shown in a lower graph 106 of FIG. 2, boundaries 107 of lanes 101 to 104 of each base are shown.
It becomes maximum. Lanes 101 to 104 are 4 for A, G, C, and T, respectively.
The sequence (base sequence) of each band 105 on the film 10 can be uniquely determined if it corresponds to one base and the boundary can be accurately detected. In the present invention, each boundary 107 of the lanes 101 to 104 can be accurately detected as each X coordinate at which the sum of absolute differences is maximized.

FIG. 1 shows a block diagram of an embodiment of the method of the present invention. This is a one-dimensional image sensor 11 for reading the image of the DNA sequence film 10, an analog / digital conversion circuit 12 for converting its output from analog to digital, and between each pixel in the X-axis direction (main scanning direction). A difference absolute value circuit 13 for obtaining an absolute value (difference absolute value) of a change amount (difference) of image data, Y for each same X coordinate
It is composed of an addition processing unit 14 for obtaining the sum of absolute values of errors in the axial direction (sum of absolute differences). The image data of the analog-digital conversion circuit 12 (image data of the band 105) and the sum of absolute difference values of the addition process 14 are input to the computer main body, and the base sequence of the gene is finally determined by digital processing.

As shown in FIG. 2, the image of the DNA sequence film 10 is scanned with the X-axis direction as the main scanning direction and the Y-axis direction as the sub-scanning direction, and is scanned by the one-dimensional image sensor 11 for each line in the X-axis direction. Read. The image data read by the image sensor 11 is converted from an analog value to a digital value in the analog / digital conversion circuit 12 for each pixel. Here, the density of one pixel is represented by an 8-bit digital value. The image data for one line converted into the digital value is input to the absolute difference value circuit 13.

FIG. 3 is a detailed diagram of the difference absolute value circuit 13, which includes a delay circuit 131 and a difference absolute value calculation circuit 132. The delay circuit 131 delays the input image data by one pixel, and the difference absolute value calculation circuit 132 outputs the original input image data which is not delayed and 1
This is a circuit that finds the absolute value of the difference from the image data delayed by the number of pixels. The image data for one line, which is sequentially sent from the analog / digital conversion circuit 12 in series, is input to the difference absolute value calculation circuit 132 and the delay circuit 1
The difference absolute value calculation circuit 13 is delayed by one pixel by 31.
It is input to 2 and the absolute value of the change between the adjacent pixels (the absolute difference value) is sequentially obtained.

FIG. 4 is a timing chart for explaining the operation of the absolute difference circuit 13. The upper part of FIG. 4 shows the input image data sent from the analog-digital conversion circuit 12, which is d _{i, j}
Indicates the data of one pixel, where the x coordinate is i and the y coordinate is j.
This input image data is passed through the delay circuit 131 to
It is delayed by the number of pixels, and the delay circuit 131 outputs the image data shown in the middle part of FIG. Difference absolute value calculation circuit 13
In 2, the absolute value of the difference between the input image data (upper data in FIG. 4) and the image data delayed by one pixel (middle data in FIG. 4) is calculated, and the absolute difference value shown in the lower part of FIG. 4 is obtained. Output the data. Here, the absolute difference value D _{i, j} is represented by D _{i, j} = | d _{i, j} −d _{i, j−1} |.

In this way, the absolute value of the change (difference absolute value) between the adjacent pixels in the X-axis direction is obtained almost in parallel with the running of the DNA sequence film 10. The difference absolute value data obtained by the difference absolute value circuit 13 is input to the addition processing unit 14, and the sum of the difference absolute values in the Y axis direction is calculated for each coordinate in the X axis direction.

FIG. 5 is a detailed diagram of the addition processing unit 14, which includes an addition base 141, an address counter 142, and a cumulative value memory 143. Each address of the cumulative value memory 143 corresponds to each coordinate in the X-axis direction, and the difference absolute value sum (partial sum) data of the coordinate calculated so far is stored for each address. The adder 141 is a circuit that adds the input difference absolute value from the difference absolute value circuit 13 and the difference absolute value up to that point in the cumulative value memory 143. The address counter 142 specifies a read / write address of the cumulative value memory 143, is sequentially counted up by a clock, and when the maximum value (maximum X coordinate value) is reached, the operation of returning to 0 is repeated. In Fig. 5, the X coordinate is 0-2.
^Since there are up to ¹³ (the number of pixels in the X-axis direction =
2 ¹³ ), and the address counter 142 has a 14-bit configuration.

The difference absolute value data corresponding to each pixel (each X coordinate) for one line obtained by the difference absolute value circuit 13 is sequentially input to the adder 141 in synchronization with the clock. The clock is input to the address counter 142 and used to generate the address of the cumulative value memory 143. The address counter 142 sequentially reads the absolute difference value sum data corresponding to each X coordinate from the cumulative value memory 143 in the same cycle as the clock, and inputs the sum data to the adder 141. The addition result of the adder 141 is rewritten to the address corresponding to the X coordinate in the cumulative value memory 143. By repeating this for each line, when all the images on the DNA sequence film 10 have been read, the cumulative value memory 143 indicates the magnitude of the image change in the X-axis direction as shown by 106 in FIG. The characteristic amount (sum of absolute differences) is obtained as a digital value for each X coordinate.

In FIG. 5, the difference absolute value sum data in the cumulative value memory 143 has a 16-bit configuration. In contrast,
The difference absolute value data obtained by the difference absolute value circuit 13 has an 8-bit structure. Therefore, in the adder 141, after adding the 8-bit all-zero data to the upper digit of each input differential absolute value data for one line, the cumulative value memory 14 is added.
It will be added to the 16-bit differential absolute value sum data read from 3.

Although not shown in FIG. 1, the sum of absolute difference values of each X coordinate of the cumulative value memory 143 in the addition processor 14 is input to the computer main body. Image data of the DNA sequence film 10 is also input to the computer main body as an output of the analog / digital conversion circuit 12, and is stored in a two-dimensional memory of an image similar to that of the film 10. In the computer body, with respect to the image data of the two-dimensional memory, the sum of absolute differences in the X-axis direction obtained by the addition processing unit 14 is referred to, and the X-coordinate at which it becomes maximum is a band (lane) corresponding to each base. It will be recognized as the boundary of and the base sequence will be determined.

In the embodiment, the absolute value of the difference value is obtained by digitizing the image data, but it is obvious that the calculation can be performed using an analog signal. Further, it is clear that using the square value of the difference or the like instead of the absolute value of the difference value has the same effect.

Although the delay circuit 131 in FIG. 3 shows a delay corresponding to one pixel, the same effect can be obtained by adding a delay of two pixels or more.

〔The invention's effect〕

As is clear from the above description, according to the present invention, it is possible to positively utilize the characteristic that the image (band) of individual bases appears alternately on the DNA sequence film in a direction perpendicular to the migration direction. It is possible to clearly detect the boundaries of bands (lanes) corresponding to each base of A, G, C, and T, and a stable recognition rate for various patterns can be obtained in the automatic base sequence recognition device. Can be secured. In addition, the characteristic quantity having a peak at the boundary of the lane corresponding to each base is
Since it is obtained almost simultaneously with the reading of the image on the DNA sequence film, little extra processing time is required to obtain the characteristic quantity.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a diagram for explaining boundary recognition between base lanes according to the present invention, and FIG.
FIG. 4 is a detailed diagram of the differential absolute value circuit in FIG. 1, FIG. 4 is a diagram showing an operation timing chart of the differential absolute value circuit, and FIG. 5 is a detailed diagram of the addition processing unit in FIG. 10 …… DNA sequence film, 11 …… One-dimensional image sensor, 12 …… Analog / digital conversion circuit, 13 …… Differential absolute value circuit, 14 …… Addition processing unit, 101 to 104 …… Lane, 105 …… Band .

Continuation of the front page (56) References JP-A 61-99841 (JP, A) JP-A 62-102375 (JP, A) JP-A 62-85861 (JP, A)

Claims (1)

[Claims] 1. An image of a base sequence on a film (hereinafter referred to as a DNA sequence film) obtained by labeling a base sequence of a gene with a radioisotope and the like and associating it with a band corresponding to each base. In the DNA pattern reader, the image of the base sequence on the DNA sequence film is sub-scanned in the electrophoresis direction (Y-axis direction) and perpendicular to the electrophoresis direction (X-axis direction). For sequentially reading as main scanning, for the read image data of each main scanning line, a means for obtaining an absolute difference value of the image data between adjacent pixels in the X-axis direction, and each obtained main scanning line Absolute difference values in the X-axis direction are added for the same X-coordinate in the Y-axis direction to obtain a plurality of X-coordinate positions where the addition result is maximum, and the obtained X-coordinate positions are adjacent to each other. DNA pattern reading apparatus characterized by having a means for a set of things the boundary of each band corresponding to each base, the.