JP7370519B2 - Prediction program, prediction method and prediction device - Google Patents

Description

The present invention relates to a prediction program, a prediction method, and a prediction device.

The growth of seedlings such as tomatoes varies greatly depending on the environment, such as the variety and the seedling raising equipment. In the past, people who were familiar with the cultivation physiology of crops used their past experience and intuition to conduct repeated cultivation tests comparing new and conventional seedling-raising environments through trial and error. Seedlings were to be cultivated in an optimal seedling-raising environment (a highly productive seedling-raising environment).

Note that Patent Documents 1, 2, and the like disclose techniques for providing plant growing information including control information of control devices and information regarding plant growing work to customers through a network.

Japanese Patent Application Publication No. 2006-333744 Japanese Patent Application Publication No. 2006-254775

The seedling growing environment set through trial and error as described above can be maintained at a certain level if the same conditions are repeated every year. However, there is no basis for setting the seedling growing environment, so if the region, environment, facilities, etc. change, or the variety changes, it is difficult to set the seedling growing environment in accordance with these changes. Therefore, if there is a change in the facility or variety, trial and error must be performed again, and it takes a long time to set the optimal seedling-raising environment.

In addition, in the current situation where a generational change is inevitable due to the aging of facility cultivation producers, there is a need for technology that can replace the knowledge and experience of people who have been conducting cultivation tests for many years.

The present invention aims to provide a prediction program, a prediction method, and a prediction device that are capable of predicting cultivation conditions for making crops have a desired grass shape, or predicting crop grass shapes according to cultivation conditions. purpose.

The present inventors have studied various means for solving the above problems. The present inventors have been conducting research to analyze the dynamics of gene expression and metabolites in plants using molecular biological methods (so-called omics research), and have accumulated a lot of knowledge with high academic value. It's coming. The present inventors have also discovered that by using these findings to elucidate the physiological mechanism from molecular fluctuations within the plant body, it is possible to construct an appropriate cultivation environment according to the variety or item in a short time. The present invention was completed based on the circumstances described above.

The prediction program of the present invention determines the relationship between the value of an environmental factor and the expression level of a gibberellin-related gene and/or a cytokinin-related gene in a crop, and the expression of a gibberellin-related gene and/or a cytokinin-related gene in the crop. A model representing the relationship between the value of the environmental factor and the value indicating the grass shape is generated based on the relationship between the amount and the value indicating the grass shape of the seedling of the crop, and the desired value of the environmental factor or Accepts input of a value indicating a desired grass appearance, and uses the model to generate a value indicating a desired grass appearance corresponding to the value of the environmental factor, or an environment corresponding to a value indicating the desired grass appearance. This is a prediction program that causes a computer to predict and output the value of a factor.

The prediction program, prediction method, and prediction device of the present invention have the effect of being able to predict the cultivation conditions for making crops have a desired grass appearance, or predict the grass appearance of crops according to the cultivation conditions.

Figure 1(a) is a photograph of a cross-section (section) of a stem of a seedling of the variety "CF Momotaro York", and Figure 1(b) is a photograph of a cross-section (section) of a stem of a seedling of the variety "Tomimaru Mucho". be. It is a graph showing the relationship between xylem area and total dry weight (value related to yield) of tomato seedlings. Figure 3(a) shows the trees of five varieties, ``Managua,'' ``Endeavor,'' ``Tomimaru Mucho,'' ``CF Momotaro York,'' and ``Reiyo,'' at 3, 4, and 5 weeks after sowing. FIG. 3(b) is a graph showing the proportion of subtissues, and FIG. 3(b) is a graph showing the amount of cytokinin in each variety three weeks after sowing. Figure 4 (a) is a graph showing the relative expression level of the cytokinin synthesis gene (SlIPT3) in the case of high nitrogen and in the case of medium nitrogen, and Figure 4 (b) is a graph showing the relative expression level of the cytokinin synthesis gene (SlIPT3) in the case of high nitrogen and in the case of medium nitrogen. 2 is a graph showing the relative expression levels of cytokinin-responsive genes in FIG. It is a figure showing the influence by nutrient solution nitrogen concentration. FIG. 6(a) is a graph showing the relationship between the nitrogen ion concentration in the nutrient solution and the expression of cytokinin-responsive genes, and FIG. 6(b) is a graph showing the relationship between the endogenous amount of cytokinin and the expression of cytokinin-responsive genes. This is a graph showing. It is a graph showing the relationship between the endogenous amount of cytokinin and the amount of xylem growth per day. This is a photograph comparing sections of hypocotyl and epicotyl of tomato seedlings grown at temperatures of 25°C and 30°C. Figure 9(a) is a diagram showing -DIF, Figure 9(b) is a diagram showing +DIF, and Figure 9(c) is a diagram showing plant heights of seedlings grown under +DIF and -DIF conditions. It's a photo. Figure 10(a) is a graph showing the expression level of the cell wall elongation enzyme gene (LeEXP2 gene) on the 14th day when seedlings were grown under +DIF and -DIF conditions, and Figure 10(b) is a graph showing the expression level of the LeEXP2 gene on the 14th day. It is a graph showing the relationship between the expression level and the endogenous amount of gibberellin. It is a graph showing the relationship between the endogenous amount of gibberellin and the amount of hypocotyl elongation per day. Figure 12(a) shows the relationship between the hypocotyl length (predicted value on day 21 after seeding) and the expression level of gibberellin-responsive genes (actual value on day 14 after seeding) when various conditions were changed. FIG. 12(b) is a graph showing the relationship between the predicted value of the hypocotyl length on the 21st day after seeding and the measured value of the hypocotyl length on the 21st day after seeding. It is a figure showing a seedling raising regression model. It is a table showing details of key genes. 1 is a diagram schematically showing the configuration of an information processing system according to an embodiment. FIG. 16(a) is a diagram showing the hardware configuration of the server, and FIG. 16(b) is a functional block diagram of the server. FIG. 3 is a diagram showing an overview of server processing. FIG. 3 is a diagram showing an example of a screen displayed on a user terminal.

Hereinafter, an information processing system according to one embodiment will be described in detail based on FIGS. 1 to 18.

The information processing system of this embodiment predicts and outputs what kind of environmental settings should be made to make tomato seedlings "good seedlings," and predicts and outputs what kind of environmental settings should be made to make tomato seedlings "good seedlings", This is a device that predicts and outputs the appearance of grass. In the present embodiment, a "good seedling" refers to a highly productive seedling, and refers to a seedling having a predetermined ideal plant appearance (stem diameter, plant height). In general, thicker stems have better material transport properties and are likely to yield higher yields. Also, shorter plants are easier to handle when transplanting tomato seedlings.

Here, before explaining the information processing system of this embodiment, the knowledge that the inventors used when constructing the information processing system will be explained.

(Knowledge 1: Regarding xylem development of crops)
The present inventors focused on the fact that there are differences between cultivars in the development of xylem tissue in stems. For example, the results of comparing the xylem tissue area (xylem area) of the tomato variety "CF Momotaro York" and the variety "Tomimaru Mucho" are shown in Figures 1(a) and 1(b). . Figure 1(a) shows a photograph of a cross-section (section) of a stem of a seedling of ``CF Momotaro York'', and Figure 1(b) shows a photograph of a seedling of ``Tomimaru Mucho'' (same as Figure 1(a)). A photograph of a cross-section (section) of a stem (seedling at the same time) is shown. In FIGS. 1(a) and 1(b), the portion indicated by double-headed arrows is the xylem tissue. As can be seen by comparing Figures 1(a) and 1(b), "Tomimaru Mucho" has a larger xylem area.

Here, "Tomimaru Mucho" is generally known to have a higher yield than "CF Momotaro York". In other words, since Tomimaru Mucho, which has a higher yield, has a larger xylem area, it is expected that the higher the yield, the larger the xylem area.

Regarding this point, we analyzed the recombinant inbred lines (RILs) of the high-yielding variety "Geronimo" and the low-yielding variety "Momotaro 8" and determined the xylem area (mm ² ) of tomato seedlings. Figure 2 shows the relationship with total dry weight (g). Generally, the greater the total dry weight, the greater the fruit weight and therefore the yield. As shown in FIG. 2, it can be seen that the larger the xylem area, the larger the total dry weight, and therefore the larger the xylem area, the larger the yield.

The present inventors also verified the relationship between xylem development and the amount of cytokinin in seedlings. Figure 3(a) shows the results of five varieties ``Managua'', ``Endeavor'', ``Tomimaru Mucho'', ``CF Momotaro York'', and ``Reiyo'' at 3, 4, and 5 weeks after sowing. The percentage of xylem tissue (the ratio of the area of xylem tissue to the entire section as shown in FIGS. 1(a) and 1(b)) is shown. Figure 3(a) shows that there is not much difference in the percentage of xylem tissue of each variety in the 3rd week after sowing, but in the 5th week after sowing, "Managua", "Endeavor", It can be seen that the percentage of xylem tissue in "Maru Mucho" increases, and the percentage of xylem tissue in "CF Momotaro York" and "Reiyo" decreases.

Further, FIG. 3(b) shows the results of examining the amount of cytokinin (iP7G: isopentenyladenine-7-glucoside) in each variety 3 weeks after sowing. As shown in Figure 3(b), 3 weeks after sowing, the amount of cytokinin was high in "Managua", "Endeavor", and "Tomimaru Mucho", and the amount of cytokinin in "CF Momotaro York" and "Reiyo" was high. It turns out that there are few.

That is, when FIGS. 3(a) and 3(b) are combined, it can be seen that the more the xylem tissue of the stem develops, the more the cytokinin amount increases before the xylem tissue starts to develop.

Furthermore, the present inventors verified the relationship between the amount of cytokinin and the nitrogen concentration of the nutrient solution when tomato was cultivated using hydroponic cultivation. Figure 4(a) shows the expression levels of the cytokinin synthesis gene (SlIPT3 gene) when the nitrogen concentration is high (90 μmol/L, high nitrogen) and when it is lower (50 μmol/L, medium nitrogen). The expression level is shown as a relative expression using an internal standard. From FIG. 4(a), it was found that under high nitrogen conditions, the number of cytokinin synthesis genes (SlIPT3 gene) increases, so that cytokinin biosynthesis increases. Note that details of the SlIPT3 gene are shown in the table of FIG. 14. Note that FIG. 14 also lists details of other genes that will be described in the following explanation.

In addition, Figure 4(b) shows the expression levels of cytokinin-responsive genes in the case of high nitrogen and medium nitrogen as relative expression using the expression level of 18S rRNA as an internal standard. There is. As shown in the graph on the left of FIG. 4(b), it was found that more cytokinin-induced genes (SlRRA6 gene, SlRRA2 gene) were produced under high nitrogen conditions than under medium nitrogen conditions. Furthermore, as shown in the graph on the right of FIG. 4(b), it was found that less cytokinin suppressive gene (SlRRB11 gene) was produced under high nitrogen conditions than under medium nitrogen conditions.

Therefore, from these relationships, the following was found as shown in FIG. That is, when the nitrogen nutrient in the nutrient solution increases (becomes highly concentrated), the number of cytokinin synthesis genes increases, resulting in an increase in cytokinin, which in turn increases the number of cytokinin-inducible genes and decreases the number of cytokinin-repressing genes. It was also found that vascular bundle development was promoted and yield increased due to an increase in cytokinin-induced genes and a decrease in cytokinin-inhibitory genes. This is considered to mean that the expression level of cytokinin-related genes changes before a difference in stem diameter appears due to the influence of the nutrient solution nitrogen concentration.

Based on the above-mentioned findings, the present inventors expressed the relationship between nitrogen ion concentration and the expression of cytokinin-responsive genes using a regression model. FIG. 6(a) shows the relationship between the nitrogen ion concentration (μmol/L) in the nutrient solution and the expression of cytokinin-responsive genes. Furthermore, the present inventors expressed the relationship between the endogenous amount of cytokinin and the expression of cytokinin-responsive genes using a regression model. FIG. 6(b) shows the relationship between the endogenous amount of cytokinin (pmol/g) and the expression of cytokinin-responsive genes. Furthermore, the present inventors expressed the relationship between the endogenous amount of cytokinin (pmol/g) and the amount of xylem growth per day (cm ² ) using a regression model. FIG. 7 shows the relationship between the endogenous amount of cytokinin and the amount of xylem growth per day.

From Figure 6(a), if the environmental measurement value (nitrogen concentration) is e and the coefficient for each variety is cv ₁ , the key gene expression level (cytokinin responsive gene expression level) _Kg can be calculated using the following formula (1). can be expressed.
K _g =f ₁ (e, cv ₁ )...(1)

Furthermore, from FIG. 6(b), if the coefficient for each variety is cv ₂ , the endogenous amount of key hormone (endogenous amount of cytokinin) K _h can be expressed by the following equation (2).
K _h = f ₂ (K _g , cv ₂ )...(2)

Further, from FIG. 7, if the coefficient for each variety is cv ₃ , the xylem growth amount z per day can be expressed by the following equation (3).
z=f ₃ (K _h , cv ₃ )...(3)

The xylem area (stem diameter) , can be expressed by the following equation (4).
X=f(e, _cv4 )...(4)

Note that cv ₄ means a coefficient reflecting the expression characteristics of a key gene (cytokinin-responsive gene) of the variety.

As shown in the above equation (4), the present inventors were able to create a model capable of simulating the xylem area (stem diameter) X from the environmental measurement value (nitrogen concentration) e.

(Finding 2: Regarding the increase in stem diameter and plant height under high temperature conditions)
The present inventors cultivated tomato seedlings of the same variety under conditions of a temperature of 25° C. and a condition of a temperature of 30° C. (high temperature conditions) under the same conditions except for the temperature. They then compared the sections of the hypocotyl and epicotyl of the cultivated tomato seedlings. FIG. 8 shows the comparison results. From FIG. 8, it can be seen that the stem diameter becomes larger under high temperature conditions. Although not shown in the figure, it was found that the plant height also increased when the temperature was 30°C.

(Finding 3: Plant height becomes smaller under day-night temperature reversal conditions)
The day/night temperature reversal condition means a case where the temperature is set low during the light period (day) and high during the dark period (night), as shown in FIG. 9(a). This day/night temperature reversal condition is called -DIF, and is a condition in which the value obtained by subtracting the dark period temperature from the light period temperature is negative. On the other hand, as shown in FIG. 9(b), when the temperature is high during the light period and low during the dark period, it is called +DIF.

The present inventors verified the plant height of tomato seedlings under +DIF conditions and -DIF conditions. As a result, as shown in FIG. 9(c), it was found that the plant height of the seedlings was lower in -DIF.

Here, the present inventors focused on the LeEXP2 gene (see Figure 14), which is a cell wall elongation enzyme gene, and determined that the cell wall elongation enzyme gene (LeEXP2 gene) on day 14 when seedlings were grown under +DIF and -DIF conditions. The expression level of was measured. This result is shown in FIG. 10(a). From FIG. 10(a), it was found that there was less LeEXP2 gene under the −DIF condition than under the +DIF condition. In addition, in the case of -DIF conditions, the hypocotyl elongation regulating transcription factor gene (SlHY5 gene), whose function is known to change depending on the light, temperature, and oxygen environment, increases, and the temperature/light signal transduction gene (SlPIF4/ 5 gene) is also reduced, it is thought that the expression of genes involved in cell wall elongation, such as the LeEXP2 gene, is reduced.

Furthermore, based on the knowledge that the cell wall elongation enzyme gene (LeEXP2 gene) is a gibberellin-responsive gene, the present inventors investigated the expression level of the LeEXP2 gene and the endogenous amount of gibberellin (GA4: active form) (pmol/FW). ) and found a relationship as shown in FIG. 10(b). As shown in FIG. 10(b), there was a roughly proportional relationship between the expression level of the LeEXP2 gene and the endogenous amount of gibberellin (GA4).

Furthermore, the present inventors verified the relationship between the endogenous amount of gibberellin (GA4) (pmol/FW) and the amount of hypocotyl elongation (cm) per day, and obtained the results shown in Figure 11. I was able to do that. From FIG. 11, there was a roughly proportional relationship between the endogenous amount of gibberellin (GA4) and the amount of hypocotyl elongation per day.

Here, from FIG. 10(a), if the environmental measurement value (-DIF) is e' and the coefficient for each variety is cv ₁ ', the key gene expression level K _g ' can be expressed by the following equation (5). I can do it. Note that the key genes in this case include the SlPer gene (see Figure 14) in addition to the LeEXP2 gene, so the key gene expression level K _g ' in the following equation (5) is calculated by combining the LeEXP2 gene and SlPer gene expression levels. means.
K _g '=f ₁ '(e', cv ₁ ')...(5)

Further, from FIG. 10(b), if the coefficient for each variety is cv ₂ ', the endogenous amount of key hormone (endogenous amount of gibberellin) K _h ' can be expressed by the following equation (6).
K _h '=f ₂ '(K _g ', cv ₂ ')...(6)

Further, from FIG. 11, if the coefficient for each breed is cv ₃ ', the amount of hypocotyl elongation z' per day can be expressed by the following equation (7).
z'=f ₃ '(K _h ', cv ₃ ')...(7)

Then, the plant height (hypocotyl length) X' increases as the amount of hypocotyl elongation z' increases per day. can be expressed by the following equation (8).
X'=f'(e', cv ₄ ')...(8)

Note that cv ₄ ' means a coefficient reflecting the expression characteristics of key genes (LeEXP2 gene, SlPer gene) of the variety.

As shown in equation (8) above, the present inventors were able to create a model that can simulate plant height (hypocotyl length) X' from the environmental measurement value (-DIF): e'.

(Finding 4: Regarding other environmental factors that affect plant height)
The present inventors also examined other environmental factors that are thought to affect plant height. For example, PBZ (paclobutrazol), a plant growth regulator, is known as a gibberellin biosynthesis inhibitor. Therefore, it is expected that by feeding PBZ to seedlings, the synthesis of gibberellin will be inhibited, resulting in a decrease in plant height. Furthermore, under high temperature conditions, the plant height increases as mentioned above. The present inventors investigated the hypocotyl length (mm) (predicted value on day 21 after seeding) and the expression level of the gibberellin-responsive gene (LeEXP2 gene) when variously changing the conditions of -DIF, PBZ, and high temperature. (Actual measurement value on the 14th day after sowing) was plotted as shown in FIG. 12(a). Note that the hypocotyl length (predicted value on the 21st day after sowing) in Figure 12(a) almost matches the hypocotyl length (actual value on the 21st day after sowing) as shown in Figure 12(b). Therefore, the predicted values in FIG. 12(a) can be regarded as actual measured values.

According to FIGS. 12(a) and 12(b), it can be seen that the expression level of the gibberellin-responsive gene (LeEXP2 gene) changes by varying the amount of PBZ given to the seedlings and the high temperature conditions. Therefore, the hypocotyl length determined from the above formula (8) needs to be corrected depending on the amount of PBZ and whether or not it is a high temperature condition.

(Other findings)
The rate of photosynthetic sugar production is affected by light and CO ₂ concentration. Photosynthetic sugar production is also affected by high temperature and nutrient solution nitrogen concentration. It is known that when a large amount of sugar is synthesized, the sugar energy-related enzyme gene (SlTPS1 gene) increases, and at this time, cytokinin also increases. Therefore, since light and CO ₂ concentration also affect the development of xylem tissue, that is, the stem diameter, the stem diameter determined from the above equation (4) needs to be corrected by taking into account the effects of light and CO ₂ concentration. There is.

(Summary of knowledge)
The above findings can be summarized as a seedling-raising regression model as shown in FIG. 13. This seedling-raising regression model is a model that expresses the relationship between the transcriptome, hormoneome, and metabolome. The T-shaped arrow extending from "PBZ" to "gibberellin response/cell elongation" and the T-shaped arrow extending from "respiration" to "gibberellin response/cell elongation" indicate genes (LeEXP2 gene, SlPer gene). This means the effect of suppressing the expression level of .

By using the seedling-raising regression model shown in Figure 13 as a machine learning inference model, if you input PBZ, -DIF, temperature, nutrient solution nitrogen concentration, and light/CO ₂ concentration as environmental factors (seedling-raising conditions), these Based on this, it is possible to predict the plant height and stem diameter of seedlings within a predetermined number of days after sowing. The predetermined number of days after sowing can be, for example, the 21st day after sowing, the 17th day after sowing, etc. in the case of self-rooted seedlings. Furthermore, by inputting a desired grass appearance (plant height, stem diameter) over a predetermined number of days into the inference model, it is possible to output the values of environmental factors that should be set to achieve the desired grass appearance.

(Information processing device)
FIG. 15 shows the configuration of an information processing system 100 according to this embodiment.

As shown in FIG. 15, the information processing system 100 includes a server 10 as a prediction device and a user terminal 70. The server 10 and the user terminal 70 are connected to a network 80 such as the Internet. The user terminal 70 is a terminal used by a user who grows tomatoes, and is a PC (Personal Computer), a tablet terminal, a smartphone, or the like.

The server 10 is an information processing device that generates the above-described inference model by machine learning and performs simulation using the generated inference model. FIG. 16(a) shows the hardware configuration of the server 10. As shown in FIG. 16A, the server 10 includes a CPU 90, a ROM 92, a RAM 94, a storage section (HDD in this case) 96, a network interface 97, a portable storage medium drive 99, and the like. Each component of the server 10 is connected to a bus 98. In the server 10, the CPU 90 executes a program (including a prediction program) stored in the ROM 92 or HDD 96, or a program (including a prediction program) read from the portable storage medium 91 by the portable storage medium drive 99. Thus, the functions of each part shown in FIG. 16(b) are realized. Note that the functions of each part in FIG. 16(b) may be realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

FIG. 16(b) shows a functional block diagram of the server 10. Further, FIG. 17 shows an overview of the processing of the server 10. As shown in FIG. 16(b), in the server 10, the CPU 90 functions as a past data reception section 20, an inference model generation section 22, an input reception section 24, and a prediction section 26 by executing a program.

The past data reception unit 20 receives a large amount of past data (varieties of seedlings cultivated in the past, environmental factors, key gene expression levels, plant height and stem diameter in a predetermined number of days) input by a system administrator, etc., and receives the input data from the inference model generation unit. Hand it over to 22nd.

The inference model generation unit 22 generates an inference model using a large amount of past data as teacher data (step S10 in FIG. 17). This inference model is a model based on Figure 13, and by inputting the variety and environmental factors, it is possible to predict and output the plant height and stem diameter in a predetermined number of days. This is a model that can predict and output appropriate environmental factors by inputting.

Further, when at least one past data of the new variety is received (step S11 in FIG. 17), the inference model generation unit 22 inputs the past data of the new variety into the inference model, thereby converting the inference model into a new variety. Adjustments are made accordingly (step S12). In this adjustment, processing is performed to determine the coefficients of equation (4) and equation (8) described above.

The input receiving unit 24 receives information input by the user (the variety and environmental factors, or the variety and plant height/stem diameter in a predetermined number of days), and passes it to the prediction unit 26 (step S14 in FIG. 17).

The prediction unit 26 uses the inference model corresponding to the variety input by the user to predict and output the plant height and stem diameter in a predetermined number of days from the environmental factors input by the user (steps S16 and S18 in FIG. 17). Furthermore, the prediction unit 26 uses the inference model corresponding to the variety input by the user to predict and output environmental factors to be set based on the plant height and stem diameter for a predetermined number of days input by the user (S16, S18). .

As described above, in this embodiment, an inference model based on FIG. Predict the environmental factors that will lead to the desired grass appearance. This makes it possible to accurately predict expected plant appearance and environmental factors to be set using the relationships among the transcriptome, hormonome, and metabolome.

FIG. 18 shows an example of a screen of an application used by the user on the user terminal 70. The screen shown in FIG. 18 is provided with columns A to E. For example, the user inputs the variety and sowing date in column A, and inputs the desired plant height and stem diameter for a predetermined number of days (for example, the 21st day) from the sowing date in column D. Note that column C is a fixed item (value is fixed). In this state, when the user presses the send button in column E, the input information is sent from the user terminal 70 to the server 10. Then, the server 10 inputs the input information into an inference model to predict the appropriate EC (nitrogen concentration), light period temperature, dark period temperature, and PBZ processing amount for each day after sowing, and the user terminal Send to 70. As a result, appropriate environmental factor information is displayed in the blank space on the screen in FIG. 18.

The user can also input the variety, sowing date, PBZ in column A of FIG. 18, EC, light period temperature, and dark period temperature in column B, and press the send button. In this case, input information is transmitted from the user terminal 70 to the server 10. The server 10 inputs the input information into an inference model to predict the plant height and stem diameter for a predetermined number of days (for example, the 21st day) after sowing, and sends the predicted plant height and stem diameter to the user terminal 70 . As a result, information on the predicted plant height and stem diameter is displayed on the screen in FIG. 18.

Note that FIG. 18 is an example, and other input fields may be provided. For example, there may be a field for inputting grafting information (scion, rootstock information), or there may be a field for inputting other cultivation conditions.

As described in detail above, according to the server 10 of the present embodiment, the inference model generation unit 22 generates a relationship between the value of an environmental factor and the expression level of a gibberellin-related gene and a cytokinin-related gene in a crop; and the relationship between the expression levels of gibberellin-related genes and cytokinin-related genes in the crop and the values indicating the plant appearance (plant height and stem diameter) of crop seedlings, and the values of the environmental factors and the values indicating the plant appearance. Generate an inference model that represents the relationship between. In addition, the input receiving unit 24 receives an input of a value of an environmental factor desired by the user or a value indicating a desired grass appearance inputted at the user terminal 70. Then, the prediction unit 26 uses the inference model to predict a value indicating the grass appearance corresponding to the desired value of the environmental factor or a value of the environmental factor corresponding to the value indicating the desired grass appearance, and predicts the value of the environmental factor corresponding to the value indicating the desired grass appearance. Output to the terminal 70. As a result, in this embodiment, it is possible to accurately predict expected plant appearance and environmental factors to be set using the relationships among the transcriptome, hormonome, and metabolome in a short time.

Further, in the present embodiment, the inference model generation unit 22 inputs data obtained when raising seedlings of a new variety of crops into the inference model, adjusts the inference model according to the new variety, and adjusts the inference model to the new variety. 26 makes a prediction using the adjusted inference model. By inputting test data of a new variety into the inference model, it is possible to adapt the inference model to the new variety, so it is possible to adjust the inference model to appropriate environmental factors when cultivating seedlings of the new variety. Grass appearance can be predicted with high accuracy.

In the above embodiment, when the user inputs a desired environmental factor, the server 10 predicts the plant height and stem diameter of a seedling corresponding to the input environmental factor, but the present invention is not limited to this. , it is also possible to predict only one of the plant height and stem diameter of the seedling. In addition, although the server 10 has described a case in which, when the user inputs the plant height and stem diameter of a desired seedling, the environmental factors corresponding to the inputted plant height and stem diameter are predicted, but the invention is not limited to this. . For example, the user may input only one of the plant height and stem diameter of the seedling, and the server 10 may predict only the corresponding environmental factors.

Note that the user terminal 70 may have at least some of the functions of the server 10 described in the above embodiment. For example, the user terminal 70 may include the prediction unit 26 shown in FIG. 16(b). In this case, the prediction unit 26 included in the user terminal 70 may use the inference model generated by the server 10, or may generate and use the inference model within the user terminal 70.

Note that the above processing functions can be realized by a computer. In that case, a program is provided that describes the processing contents of the functions that the processing device should have. By executing the program on the computer, the above processing functions are realized on the computer. A program describing the processing contents can be recorded on a computer-readable storage medium (excluding carrier waves).

When a program is distributed, it is sold in the form of a portable storage medium such as a DVD (Digital Versatile Disc) or a CD-ROM (Compact Disc Read Only Memory) on which the program is recorded. It is also possible to store the program in the storage device of the server computer and transfer the program from the server computer to another computer via a network.

A computer that executes a program stores, for example, a program recorded on a portable storage medium or a program transferred from a server computer in its own storage device. Then, the computer reads the program from its own storage device and executes processing according to the program. Note that a computer can also directly read a program from a portable storage medium and execute processing according to the program. Furthermore, each time a program is transferred from the server computer, the computer can sequentially execute processing according to the received program.

The embodiments described above are examples of preferred implementations of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the gist of the present invention.

22 Inference model generation unit (generation unit)
24 Input reception unit 26 Prediction unit 10 Server (prediction device)

Claims (7)

The relationship between the value of an environmental factor and the expression level of a gibberellin-related gene and/or a cytokinin-related gene in a crop, and the relationship between the expression level of a gibberellin-related gene and/or a cytokinin-related gene in the crop and the seedling of the crop. generating a model representing the relationship between the value of the environmental factor and the value indicating the grass appearance, based on the relationship with the value indicating the grass appearance;
Accepting input of a desired value of an environmental factor or a value indicating a desired grass appearance,
Using the model, predicting and outputting a value indicating a grass appearance corresponding to a desired value of the environmental factor, or a value of an environmental factor corresponding to a desired value indicating the grass appearance;
A prediction program that allows a computer to perform processing. The prediction program according to claim 1, wherein the model is generated based on a relationship between the expression level of the gibberellin-related gene and a value indicating the plant height of a seedling of the crop. 3. The prediction program according to claim 1, wherein the model is generated based on a relationship between the expression level of the cytokinin-related gene and a value indicating a stem diameter of a seedling of the crop. In the generation process, data obtained when raising seedlings of a specific variety of crops is input into the model, and the model is adjusted to match the specific variety;
In the outputting process, the adjusted model is used to generate a value indicating the grass appearance of the seedling of the specific variety of crop corresponding to the desired value of the environmental factor, or a desired value of the seedling of the specific variety of crop. 4. The prediction program according to claim 1, wherein the prediction program predicts and outputs a value of an environmental factor corresponding to a value indicating a grass appearance. The prediction program according to any one of claims 1 to 4, wherein the crop is a tomato. The relationship between the value of an environmental factor and the expression level of a gibberellin-related gene and/or a cytokinin-related gene in a crop, and the relationship between the expression level of a gibberellin-related gene and/or a cytokinin-related gene in the crop and the seedling of the crop. generating a model representing the relationship between the value of the environmental factor and the value indicating the grass appearance, based on the relationship with the value indicating the grass appearance;
Accepting input of a desired value of an environmental factor or a value indicating a desired grass appearance,
Using the model, predicting and outputting a value indicating a grass appearance corresponding to a desired value of the environmental factor, or a value of an environmental factor corresponding to a desired value indicating the grass appearance;
A prediction method characterized in that processing is performed by a computer. The relationship between the value of an environmental factor and the expression level of a gibberellin-related gene and/or a cytokinin-related gene in a crop, and the relationship between the expression level of a gibberellin-related gene and/or a cytokinin-related gene in the crop and the seedling of the crop. a generation unit that generates a model representing the relationship between the value of the environmental factor and the value indicating the grass appearance, based on the relationship with the value indicating the grass appearance;
an input reception unit that receives input of a desired value of an environmental factor or a value indicating a desired grass appearance;
a prediction unit that uses the model to predict and output a value indicating a grass appearance corresponding to a desired value of the environmental factor, or a value of the environmental factor corresponding to a desired value indicating the grass appearance;
A prediction device comprising:

Images