JP2019095819A - Information processing device and program - Google Patents

Abstract

To provide an information processing device and a program with which it is possible to suppress homologous recombination and increase the production amount of target protein.SOLUTION: First generation parent individual group data is acquired that is generated on the basis of data representing an amino acid sequence, the number of genes and a codon frequency table and includes a predetermined number of individual data. A mutation process is executed on individuals included in the first generation parent individual group data, and first generation child individual group data is acquired that includes the individuals having undergone the mutation process. A non-dominance sort process is executed on first generation integrated data having been integrated from the first generation parent individual group data and the first generation child individual group data on the basis of a predetermined evaluation standard that pertains to codon suitability and the base sequence of the codon, and the entire individual data included in the first generation integrated data is classified into ranks in the Pareto optimal solution. A predetermined number of individual data is selected in descending order of rank from the entire individual data classified into ranks.SELECTED DRAWING: Figure 8

Description

  The present invention relates to an information processing apparatus and program that can increase the production amount of a target protein.

  When producing a target protein in a microorganism or the like, a method is known in which a plurality of genes encoding the target protein are introduced. Such genes are often used having the same DNA sequence. However, when a plurality of genes having the same DNA sequence are introduced, homologous recombination occurs between these genes and a part of the gene is deleted. Here, the homologous recombination is a recombination that occurs at a site where the nucleotide sequences of DNAs closely resemble (homologous site). A conceptual representation of this is shown in FIG. FIG. 1 (a) shows an example in which five genes having the same DNA sequence are introduced. When homologous recombination occurs in the second half of the second gene to the first half of the fifth gene among the five genes, as shown in FIG. 1 (b), the number of genes is up to two. It will decrease and the production efficiency of the target protein will decrease.

  Patent Document 1 discloses a method for obtaining a synthetic nucleic acid molecule, which comprises the steps of: (i) providing an amino acid sequence derived from an amino acid repeat region of a polypeptide; (ii) a plurality of samples each encoding the amino acid sequence Predicting a codon-optimized nucleic acid sequence; (iii) aligning the plurality of sample codon-optimized nucleic acid sequences by sequence homology and constructing a neighboring binding tree comprising the plurality of sample codon-optimized nucleic acid sequences; (Iv) selecting a single sample of said plurality of sample codon optimized nucleic acid sequences; and (v) obtaining a nucleic acid molecule comprising said selected sample codon optimized nucleic acid sequences It is done.

JP-A-2015-524658

  The present invention provides an information processor and program capable of suppressing homologous recombination and enhancing the production amount of a target protein.

  According to the present invention, data generated based on data representing an amino acid sequence, gene number and codon frequency table is a data representing a first generation parent population including a predetermined number of individual data. A parent individual population data acquisition unit for acquiring generation parent individual population data; a mutation processing unit for performing mutation processing on the individuals included in the first generation parent individual population data; and an individual for which the mutation processing is performed A child population data acquisition unit for acquiring a first generation child population data representing a first generation child population including the first generation, and a predetermined evaluation standard, which is an evaluation standard regarding codon suitability and the base sequence of the codon And performing non-dominant sorting on the first generation integrated data obtained by integrating the first generation parent individual population data and the first generation child individual population data, and the data is included in the first generation integrated data. Select the individual data of a predetermined number in the descending order of the rank from the non-dominated sort execution unit which classifies all individual data according to rank in the Pareto optimal solution and all individual data classified according to the rank An information processing apparatus having an individual selection unit is provided.

  According to the present invention, it is possible to suppress homologous recombination and increase the production amount of a target protein based on two different evaluation criteria.

Hereinafter, various embodiments of the present invention will be illustrated. The embodiments shown below can be combined with one another.
Preferably, when selecting the predetermined number of pieces of the individual data, the individual selecting unit sequentially selects pieces of the individual data having the same rank in descending order of congestion distance.
Preferably, the parent individual population data acquisition unit sets the individual data selected by the individual selection unit as second generation parent individual population data representing a second generation parent individual population, and the mutation processing unit, The processing by the superior sort execution unit and the individual selection unit is executed until the number of generations determined in advance is reached.
Preferably, the evaluation criteria related to the degree of codon suitability are based on the minimum value of the codon matching index of CDS which is a base sequence possessed by each individual and is a target of amino acid translation.
Preferably, the larger the minimum value of the codon matching index included in the individual, the higher the evaluation of the individual.
Preferably, the evaluation criteria for the base sequences of the codons are based on the minimum value of the number of unmatched bases representing the number of unmatched bases among the two CDSs contained in each individual.
Preferably, the larger the minimum value of the number of unmatched bases, the higher the evaluation of the individual.
Preferably, the evaluation criteria for the base sequences of the codons are the longest base sequence among the base sequences among the CDSs contained in each individual, which continuously match at different sites between each CDS or within one CDS. Based on the length of the longest common character string.
Preferably, the shorter the length of the longest common character string, the higher the value of the individual.
Preferably, the mutation processing unit performs a second mutation different from the first mutation processing and the first mutation processing on each individual data included in the g-th generation parent individual population data representing the g-th generation parent individual population. Execute the process
Preferably, the mutation processing unit substitutes the codons contained in the CDS with codons more frequently than the codons with a predetermined probability for all the CDSs contained in each individual. Run.
Preferably, the mutation processing unit is a longest common character, which is the longest base sequence among base sequences successively matched among different CDSs or in different sites within one CDS among the CDSs contained in each individual. The second mutation process is performed to replace the codon overlapping with the other codon with another codon with a predetermined probability.
Preferably, the first mutation treatment or the second mutation treatment is randomly selected.
Preferably, a cross processing unit for executing cross processing on individuals included in the first generation parent individual population data, the cross processing including a g generation parent individual population representing a g generation parent individual population A predetermined number of individual data items are extracted from the data, two individual data items are selected from the extracted individual data items, and a cross process is performed on the selected two individual data items.
Preferably, the intersection processing unit determines an intersection point from the boundaries of the codons included in the CDS included in the first individual data and the second individual data which are the selected two individual data, and the intersection point is determined. The codons included in the first individual data and the second individual data are interchanged at points.
Preferably, the computer is data generated based on data representing an amino acid sequence, gene number and codon frequency table, and a first generation parent population including a predetermined number of individual data A parent individual population data acquisition unit for acquiring generation parent individual population data, a mutation processing unit for performing mutation processing on an individual included in the first generation parent individual population data, a target including the individual on which the mutation processing has been performed A child individual population data acquisition unit for acquiring first generation child individual population data representing one generation child individual population, a predetermined evaluation criterion, which is based on the evaluation criteria regarding codon suitability and the nucleotide sequence of the codon Performing a non-dominated sort process on first generation integrated data obtained by integrating the first generation parent individual population data and the first generation child individual population data, and the first generation integrated data The non-dominated sort execution unit which classifies all the contained individual data according to the rank in the Pareto optimal solution, and selects the predetermined number of pieces of the individual data from the total individual data classified according to the rank in descending order of the rank An information processing program for functioning as an individual selection unit is provided.

When producing a target protein in microorganisms etc., it is a conceptual diagram showing the conventional method which introduce | transduces two or more genes which encode target protein, and (a) is an example which introduce | transduced five genes which have the same DNA sequence, (B) shows the result that homologous recombination occurs in the first half of the second half of the second gene to the first half of the fifth gene among the five genes, and the number of genes is reduced to two. It is a conceptual diagram showing gene sequence design concerning one embodiment of the present invention, (a) is a mode which inputs input data into the algorithm concerning the present invention, and outputs a gene sequence as output data, (b) introduces. It shows that the homologous recombination does not occur in the 5 gene data obtained, and the target protein is produced from all the genes. FIG. 2 is a diagram illustrating an example of a hardware configuration of the information processing device 1; It is an exemplary functional block diagram of information processor 1 concerning one embodiment of the present invention. It is a figure for demonstrating congestion distance, (a) is a conceptual diagram of congestion distance, (b) represents the calculation formula of congestion distance. It is a figure showing an example of the flow chart for carrying out gene sequence design concerning one embodiment of the present invention. Such processing is executed prior to the main routine shown in FIG. It is a figure showing the example of individual data. In the present embodiment, one individual is expressed as a plurality of protein coding regions (CDS) encoding the same amino acid. It is a figure showing an example of the flow chart for carrying out gene sequence design concerning one embodiment of the present invention. In S22, the cross process is optional and can be omitted as necessary. FIG. 6 is a diagram illustrating an example of a flowchart for performing intersection processing according to an embodiment of the present invention. It is a conceptual diagram showing the intersection process which concerns on one Embodiment of this invention. It is a figure which shows an example of the flowchart for implementing the mutation process which concerns on one Embodiment of this invention. It is a conceptual diagram showing the mutation process which concerns on one Embodiment of this invention, (a) is a 1st mutation process, (b) is a conceptual diagram showing a 2nd mutation process. It is a figure for demonstrating "the number of unmatched bases" which is an evaluation standard regarding the base sequence of the codon concerning one embodiment of the present invention. In the example of FIG. 13, the number of unmatched bases is five. It is a figure for demonstrating the "longest common character string" which is an evaluation criteria regarding the base sequence of the codon concerning one embodiment of the present invention. In FIG. 14, the underlined part of the solid line represents the longest common character string, and the underlined part of the broken line represents the common character string. It is a figure showing the process result in the Example of this invention. In FIG. 15, evaluation criteria for codon suitability (minimum value of codon suitability index (CAI) of CDS, which is a nucleotide sequence possessed by each individual and which is a target of amino acid translation) are shown on the horizontal axis. It is the graph which plotted the calculation result of the 1st generation, the 10th generation, and the 250th generation, respectively, with the evaluation standard (minimum value of the number of unmatched bases) about the base sequence as the vertical axis. It is a figure showing the process result in the Example of this invention. In FIG. 16, evaluation criteria for codon suitability (minimum value of codon suitability index (CAI) of CDS, which is a base sequence possessed by each individual and which is a target of amino acid translation) are shown on the horizontal axis. It is the graph which plotted the calculation result of the 1st generation, the 10th generation, and the 250th generation, respectively, with the evaluation standard (the length of the longest common character string) about the base sequence as the vertical axis.

Embodiment
Hereinafter, embodiments of the present invention will be described using the drawings. The various features shown in the embodiments described below can be combined with one another.

<Gene Sequence Design According to One Embodiment of the Present Invention>
FIG. 2 is a conceptual diagram showing gene sequence design according to an embodiment of the present invention. The gene sequence design according to one embodiment is to increase gene yield of a target protein by designing gene sequence groups that do not induce homologous recombination and introducing into a microorganism or the like. As shown in FIG. 2A, data representing a target protein and data representing N genes are used as input data, and calculation processing is performed based on an algorithm (hereinafter referred to as the present algorithm), and the calculation results Output data representing a group of N gene sequences encoding a protein. Here, the present algorithm utilizes a multi-objective genetic algorithm aiming to simultaneously optimize a plurality of objective functions having phase conflicting evaluation criteria. Thereby, as shown in FIG. 2 (b), for example, when there are five introduced genes, the homologous recombination does not occur in the five gene data, and the target protein is produced from all the genes. Here, in FIG. 2 (b), the genes of 1 to 5 are genes having different base sequences from each other.

<Hardware configuration>
Next, an example of a hardware configuration of the information processing apparatus 1 according to an embodiment of the present invention will be described with reference to FIG. The information processing apparatus 1 includes a processing unit 10, a storage unit 20, an operation unit 30, a display unit 40, and a communication unit 50. The processing unit 10 executes various arithmetic processing, and is configured of, for example, a CPU or the like. The storage unit 20 stores various data and programs, and is configured of, for example, a memory, an HDD, or an SSD. Here, the program may be pre-installed when the information processing apparatus 1 is shipped, may be downloaded as an application from a site on the Web, and may be transferred from another information processing apparatus by wireless communication. The operation unit 30 operates the information processing apparatus 1 and is configured of, for example, a touch panel, a keyboard, a voice input unit, a motion recognition apparatus using a camera, and the like. The display unit 40 displays various images (including still images and moving images), and includes, for example, a touch panel display, an organic EL display, an electronic paper, and other displays. The communication unit 50 transmits and receives various data to and from other information processing apparatuses, and is configured by an arbitrary I / O. The bus 100 is configured by a serial bus, a parallel bus, and the like, and electrically connects the respective units to enable transmission and reception of various data.

<Function block diagram>
Next, the function of the information processing apparatus 1 will be described using the functional block diagram of FIG. 4. The information processing apparatus 1 is, for example, a multifunction information terminal, and is a PC, a server, a smartphone, a tablet terminal, a smart watch, or the like. The information processing apparatus 1 includes an operation unit 30, a display unit 40, a communication unit 50, a processing unit 10, and a storage unit 20. The processing unit 10 includes an individual generation unit 101, a parent individual group data acquisition unit 102, a child individual group data acquisition unit 103, a cross processing unit 104, a mutation processing unit 105, a non-dominated sort execution unit 106, and an individual selection unit 107. In addition, the storage unit 20 includes an amino acid sequence data storage unit 201, a gene number data storage unit 202, a codon frequency table data storage unit 203, a calculation data storage unit 204, and an evaluation criteria storage unit 205.

  For the functions of the operation unit 30, the display unit 40, and the communication unit 50, refer to the description of FIG.

<Processing unit 10>
Next, the function of the processing unit 10 will be described. The individual generation unit 101 acquires data representing the amino acid sequence, the number of genes and the codon frequency table from the amino acid sequence data storage unit 201, the gene number data storage unit 202 and the codon frequency table data storage unit 203, respectively, Is used to generate p pieces of individual data representing randomly generated individuals under the restriction of coding. Here, p is a positive number parameter and can be any number.

  The parent individual population data acquisition unit 102 is data for using the p individual data generated by the individual generation unit 101 for the present algorithm, and as g-th generation parent individual population data representing a g-th generation parent individual population get. Further, the calculation in the present algorithm is to execute loop calculation in which a predetermined flow is repeatedly executed a plurality of times as described later, and the parent individual group data acquisition unit 102 generates a first generation, a second generation,. First generation parent individual population data representing a g-th generation parent individual population, second generation parent individual population data... G-th generation parent individual population data is acquired. Here, g is a positive number and represents the number of loops of calculation in the present algorithm.

  The cross processing unit 104 extracts e (predetermined even number) pieces of individual data from the g-th generation parent individual population data, selects two pieces of individual data from the extracted individual data, and selects 2 Cross processing is performed on individual data items. Then, two pieces of individual data are selected from among pieces of individual data for which cross processing has not been performed yet, and cross processing is performed. This process is repeated for all of the e individual data extracted. Specifically, the intersection point is determined from the boundaries of the codons contained in the CDS included in the first individual data and the second individual data which are the two selected individual data, and the first individual data is made bordering on the intersection point And the codons contained in the second individual data are replaced. Here, selection of two pieces of individual data is performed at random using, for example, a random number table or the like. Here, e is the largest even number not exceeding “p × Pc”. Pc is a parameter, and can be any value greater than 0 and less than 1. The method of extracting e individual data from the g generation parent individual population data including p individual data is not particularly limited, but for example, the “binary tournament selection method” can be used.

  The mutation processing unit 105 executes mutation processing on all the individuals included in the g-th generation parent individual population data. In this embodiment, a second mutation process different from the first mutation process and the first mutation process is performed on each individual data included in the g-th generation parent individual population data. Specifically, the first mutation treatment and the second mutation treatment are randomly determined for each individual data. Then, when it is determined that the first mutation treatment is performed, codons included in the CDS are replaced with codons more frequently than such codons with a predetermined probability Pm for all the CDSs included in each individual data. In addition, when it is determined to be the second mutation treatment, among the CDSs contained in each individual data, it is the longest base sequence among the base sequences which continuously coincide at different sites between each CDS or within one CDS. The codon overlapping with the longest common letter is replaced with another codon with a predetermined probability Pm. Details of these processes will be described later.

  The offspring population data acquisition unit 103 acquires g generation offspring population data representing a offspring population of the g generation including the individuals on which the mutation processing by the mutation processing unit 105 has been performed. Here, the number of individual data included in the g-th generation child individual population data is equal to the number of individual data included in the g-th generation parent individual population data, and is p. This is because the mutation processing by the mutation processing unit 105 is performed on all the individuals included in the g generation parent individual population data.

  The non-dominant sort execution unit 106 integrates the g-th generation parent individual population data and the g-th child individual population data based on a predetermined evaluation criterion based on the degree of codon suitability and the base sequence of the codon. A non-dominated sort process is performed on the g-th generation integrated data. The number of individual data included in the g-th generation integrated data is 2p (= p + p). Then, all individuals are classified for each front (per rank) in the Pareto optimal solution.

  The individual selecting unit 107 selects the individual data of the number determined in descending order of rank from the g-th generation integrated data classified for each front (per rank) in the Pareto optimal solution. For example, p can be adopted as a predetermined number. Then, the parent individual group data acquisition unit 102 acquires p pieces of individual data selected by the individual selection unit 107 as g + 1th generation parent individual group data representing a parent individual group of the (g + 1) th generation.

  Here, when selecting a predetermined number of pieces of individual data, if there is individual data of the same rank, the individual selecting unit 107 may select in order from the one with the largest crowding distance (Crowding Distance). Good. Here, the congestion distance is an average distance between two solutions on both sides of a certain solution. FIG. 5 (a) schematically shows this. Then, the congestion distance is calculated by the calculation formula of FIG. 5 (b). Here, the crowded distance corresponds to the average of the lengths of the peripheries of the squares shown by broken lines in FIG. 5 (a).

<Storage unit 20>
Next, the function of the storage unit 20 will be described. The amino acid sequence data storage unit 201 stores data representing an amino acid sequence. An amino acid sequence is intended to represent the sequence of amino acids in a protein.

  The gene number data storage unit 202 stores data representing the number of genes. Here, in the present embodiment, the number of genes represents the number of CDSs contained in individual data.

  The codon frequency table data storage unit 203 stores data representing a codon frequency table. The codon frequency table is a table summarizing the frequency of use of codons in host cells.

  The calculation data storage unit 204 includes an individual generation unit 101, a parent individual group data acquisition unit 102, a child individual group data acquisition unit 103, a cross processing unit 104, a mutation processing unit 105, a non-dominated sort execution unit 106, an individual selection unit 107, etc. It stores the calculation results in various processes by.

  The evaluation criteria storage unit 205 is a predetermined evaluation criteria, and stores evaluation criteria regarding the degree of codon suitability and the base sequence of codons. Specifically, the evaluation criteria for the degree of codon suitability are based on the minimum value of the codon suitability index of CDS which is a base sequence possessed by each individual and is a target of amino acid translation. Hereinafter, such criteria are referred to as first evaluation criteria. In the first evaluation criterion, the higher the minimum value of the codon matching index included in an individual, the higher the individual is evaluated. The first of the evaluation criteria for the base sequences of codons is based on the minimum value of the number of unmatched bases representing the number of unmatched bases out of two CDSs contained in each individual. Hereinafter, such a criterion is referred to as a second evaluation criterion. In the second evaluation criteria, individuals are evaluated higher as the minimum value of the number of unmatched bases is larger. The second of the evaluation criteria for base sequences of codons is the longest base sequence among the base sequences consecutively matched among different CDSs or within one CDS among different CDSs contained in each individual. Based on the length of the longest common character string. Hereinafter, such a criterion is referred to as a third evaluation criterion. In the third evaluation criterion, the shorter the longest common character string, the higher the individual is evaluated.

  Next, details of the various functions, processes, and criteria described above will be described with reference to FIGS.

<Pre-processing>
FIG. 6 is a diagram showing an example of a flow chart for performing gene sequence design according to an embodiment of the present invention. The process shown in FIG. 6 is a process executed prior to the main routine shown in FIG. Hereinafter, the process shown in FIG. 6 is referred to as pre-processing.

  First, in S11, the processing unit 10 acquires amino acid sequence data and gene number data from the amino acid sequence data storage unit 201 and the gene number data storage unit 202. Then, the data is stored in a storage unit such as a cache memory (not shown).

  Next, in S12, the processing unit 10 acquires codon frequency table data from the codon frequency table data storage unit 203. Then, the data is stored in a storage unit such as a cache memory (not shown).

  Next, in S13, the individual generation unit 101 generates p individual data representing an individual randomly generated under the restriction of encoding the same amino acid sequence of the protein. For example, 100 pieces of random individual data may be generated.

(Individual data)
Here, individual data will be described with reference to FIG. As shown in FIG. 7, in the present embodiment, one individual is expressed as a plurality of protein coding regions (CDS) encoding the same amino acid. In the individual data shown in FIG. 7, there are x CDSs. This is the number of genes represented by the gene number data acquired by the processing unit 10 from the gene number data storage unit 202 in S11 of FIG. Each CDS encodes the same amino acid. Here, G, I, V, E and Q shown in FIG. 7 are the amino acid sequences represented by the amino acid sequence data acquired by the processing unit 10 from the amino acid sequence data storage unit 201 in S11 of FIG. In addition, each CDS has a different base sequence.

  Returning to FIG. 6, the pre-processing will be further described. In S14, the parent individual group data acquisition unit 102 acquires p individual data randomly generated by the individual generation unit 101 in S13 as first generation parent individual group data representing a first generation parent individual group. The parent individual population data is an archive population stored in the processing in the present algorithm. And if 1st generation parent individual population data is acquired, pre-processing will be ended.

<Main routine>
Next, the main routine in the present algorithm will be described using FIG. First, in S20, the processing unit 10 sets a variable g to one. Here, g is a code representing a parent individual group of the g generation. g takes values from 1 to G (predetermined number of generations G described later).

  Next, in S21, the processing unit 10 acquires first generation parent individual population data from the parent individual population data acquisition unit 102.

  Next, in S22, the crossover processing unit 104 and the mutation processing unit 105 execute crossover processing and mutation processing on the individual data included in the first generation parent individual population data. Note that the cross process is optional and can be omitted as needed. The cross processing and mutation processing will be described below with reference to FIGS. 9 to 12.

<Crossing process>
First, the intersection processing will be described using FIGS. 9 and 10. FIG. 9 is a diagram illustrating an example of a flowchart for performing intersection processing according to an embodiment of the present invention. First, in S321, the intersection processing unit 104 sets a variable i to zero.

  Next, in S322, the cross processing unit 104 receives the g-th generation parent individual group data (the first generation parent individual when g = 1 in the main routine in FIG. 8) from the processing unit 10 or the parent individual group data acquisition unit 102. Collect data). Then, (e−i) pieces of individual data (e pieces for i = 0 at this time) are randomly extracted from p pieces of individual data included in the g-th generation parent individual population data. Although the method to be used for such extraction is not particularly limited, for example, the “binary tournament selection method” can be used. Here, e is the largest even number not exceeding “p × Pc”. Pc is a parameter, and can be any value greater than 0 and less than 1.

  Next, in S323, the intersection processing unit 104 randomly selects two pieces of individual data from the (ei) pieces of individual data. Selection of two pieces of individual data is performed at random using, for example, a random number table or the like.

  Next, in S324, the cross processing unit 104 performs cross processing on the two individual data selected in S323. Here, the cross processing will be specifically described with reference to FIG.

  As shown in FIG. 10, the two pieces of individual data selected in S323 are set as first individual data and second individual data, respectively. In the example of FIG. 10, the first individual data and the second individual data each have three CDSs and different base sequences. From these individual data, crossing points are determined. The crossover point is selected at one position from the codon-codon boundary. Such decisions may be made randomly. In this embodiment, the intersection points in the first individual data and the second individual data are the same place. Then, the codons included in the first individual data and the second individual data are interchanged at the intersection point. In the present embodiment, such processing is called cross processing.

  Returning to FIG. 9, the cross process will be further described. In S325, the intersection processing unit 104 increases the variable i by two.

  Next, in S326, the intersection processing unit 104 determines whether or not the variable i = e. Then, if the determination result is NO, the process returns to S323 again. On the other hand, if the determination result is YES, the intersection processing is ended, and the calculation result is output to the calculation data storage unit 204. Here, if it is i = 2 at present, and e is larger than 2, it returns from S326 to S323. Then, two pieces of individual data are randomly selected from (e-2) pieces of individual data for which the cross process has not been performed yet. This process is repeated until the determination result in S326 is YES, that is, the cross process is performed on all e individual data. As described above, such cross processing is optional and can be omitted as necessary.

<Mutation processing>
Next, mutation processing will be described using FIGS. 11 and 12. The mutation process is performed on all the individuals included in the g generation parent individual population data. Here, when the cross process is not performed in S22, the mutation process is performed on p individual data included in the g-th generation. On the other hand, when the cross process is executed in S22, a total of p individual data obtained by combining e individual data subjected to the cross process and p-e individual data not subjected to the cross process Perform mutation processing on

  FIG. 11 is a diagram showing an example of a flowchart for carrying out a mutation process according to an embodiment of the present invention. First, in S221, the mutation processing unit 105 randomly determines which of the first mutation processing and the second mutation processing is to be performed on each individual data included in the g-th generation parent individual population data. In this embodiment, mutation processing is performed on all p individual data included in the g-th generation parent individual population data. Here, the second mutation treatment is a mutation treatment different from the first mutation treatment.

  Next, in S222, the mutation processing unit 105 determines whether or not the determination result in S221 is the first mutation processing. Then, if the determination result is YES, the process proceeds to S223a to execute the first mutation process. On the other hand, if the determination result is NO, the process proceeds to S223b to execute the second mutation process.

(First mutation treatment)
Next, in S223a, the mutation processing unit 105 executes the first mutation processing on the individual data. Specifically, for every CDS contained in the individual data, each codon is replaced with a codon having a frequency higher than that of such a codon with a predetermined probability Pm. Here, more frequent codons are obtained from the codon frequency table data acquired from the codon frequency table data storage unit 203 in S12 in the pre-processing of FIG. Here, the first mutation process will be described with reference to FIG.

  FIG. 12 (a) shows an example in which three CDSs are included in individual data. As shown in FIG. 12 (a), in the first mutation processing, mutation processing is performed with probability Pm for all codons (5 × 3 = 15 codons) for three CDSs included in the individual data. Do. The broken line in FIG. 12A indicates the range of codons to be subjected to mutation processing with probability Pm. As an example, "GGC", which is the first codon included in the third CDS, CDS-3, is replaced with a codon more frequent than GGC with probability Pm. Here, more frequent codons are obtained from the codon frequency table data. In the example of FIG. 12 (a), “GGT” and “GGA” exist as codons more frequently than “GGC”. Thus, when there are a plurality of more frequent codons, any one codon is randomly selected and replaced with "GGC". In the absence of codons more frequently than "GGC", such substitution is not performed. Such substitutions are performed for all codons contained in the individual data. In the present embodiment, such a process is referred to as a first mutation process. Here, the first mutation treatment is intended to increase the minimum CAI value according to the first evaluation criteria described later.

(Second mutation treatment)
Returning to FIG. 11, the mutation process will be further described. In S223b, the mutation processing unit 105 executes the second mutation processing on the individual data. Specifically, among the CDSs included in the individual data, codons overlapping with the longest common character, which is the longest base sequence among the base sequences which continuously match at different sites between each CDS or within one CDS, Replace with another codon with a predetermined probability Pm. Here, the longest common character string will be described with reference to FIG.

"Longest common string"
The individual data shown in FIG. 14 includes three CDSs as an example. Here, a character string (three bases (= characters) × 5 = 15 characters) representing five codons contained in each CDS and a character string contained at another site inside another CDS or one CDS In contrast, the longest of the continuously matching strings is called the longest common string. In the example of FIG. 14, “GGCATCGTCGA” (portion underlined with a solid line) is the longest common character string, and its length (number of characters) is 11. Although "GTCGAGCAG" (a portion underlined with a broken line) is also a common character string, it has a length of 9 and is not the longest, so it can not be the longest common character string. The longest common string corresponds to a concept called the longest common substring in computer science.

  FIG. 12 (b) shows an example in which three pieces of CDS are included in individual data. As shown in FIG. 12 (b), in the second mutation process, a string (three bases (= letters) × 5) representing five codons included in the CDS for three CDSs included in the individual data. The mutation process is executed with the probability Pm on the codon overlapping with the longest common character string among = 15 characters). Here, in the example of FIG. 12 (b), the longest common character string is “GGCATCGTCGA” (portion underlined with a solid line). In the example of FIG. 12 (b), among the codons contained in the second and third CDSs, CDS-2 and CDS-3, the first to fourth codons overlap with the longest common character string. is there. The broken line in FIG. 12B indicates the range of codons to be subjected to mutation processing with probability Pm. As an example, the first codon "GGC" contained in CDS-3 is replaced with another codon with probability Pm. In the example of FIG. 12 (b), "GGT", "GGA" and "GGG" are present as codons different from "GGC". Thus, when there are a plurality of other codons, one of the codons is randomly selected and replaced with "GGC". In the case where no codon other than "GGC" exists, such substitution is not made. For example, substitution may not be possible when there is only one type of codon encoding a specific amino acid. Such substitutions are performed for those codons that overlap with the longest common string. In the present embodiment, such processing is called second mutation processing. Here, the second mutation processing is intended to increase the number of unmatched bases related to the second evaluation criteria described later, and to reduce the longest common character string.

  Then, when the first mutation processing and the second mutation processing are completed, the calculation result is output to the calculation data storage unit 204.

  Returning to FIG. 8, the main routine will be further described. In S22, mutation processing by the mutation processing unit 105 is performed on the g-th generation parent individual population data, and if necessary, cross processing by the cross processing unit 104 is performed, and then the process proceeds to S23.

  Next, in S23, the offspring population data acquisition unit 103 generates g-th generation offspring population data. Hereinafter, how to generate the g-th generation individual population data will be described for each execution of the cross process.

1. When only mutation processing is performed in S22, the offspring individual population data acquisition unit 103 newly generates p individual data on which all p individual data included in the g-th generation parent individual population data are mutated. Let g generation child individual population data.

2. When mutation processing and crossover processing are performed in S22, the offspring population data acquisition unit 103 selects e pieces of data on which crossover processing has been performed among the p pieces of individual data included in the g-th generation parent population data. In addition, p individual data in which a total of p individual data obtained by combining the p-e individual data for which cross processing has not been performed are all mutated is newly defined as the g generation child individual population data.

  Next, in S24, the processing unit 10 integrates the g-th generation parent individual population data and the g-th generation child individual population data to generate the g-th generation integrated data. As a result, the g-th integrated data includes 2p individual data.

  Next, in S25, the non-dominated sort executing unit 106 is non-dominated to the g-th generation integrated data, which is a predetermined evaluation criterion based on the degree of codon suitability and the evaluation criteria regarding the base sequences of the codons. Perform a sort Then, 2p individual data are classified for each front (per rank) in the Pareto optimal solution.

  Next, in S26, the individual selecting unit 107 selects the individual data of the number determined in descending order of rank from the g-th generation integrated data classified for each front (per rank) in the Pareto optimal solution. When selecting a predetermined number of pieces of individual data, the individual selecting unit 107 may select pieces of data in order of decreasing congestion distance when there is individual data of the same rank. Here, p can be adopted as a predetermined number. Then, the parent individual group data acquisition unit 102 generates and acquires p individual data selected by the individual selection unit 107 as g + 1th generation parent individual group data representing a parent individual group of the (g + 1) th generation. Hereinafter, the evaluation criteria will be described with reference to FIGS. 13 and 14.

<Evaluation criteria>
In this embodiment, when selecting p individual data from 2p individual data subjected to non-dominated sorting, evaluation criteria of two viewpoints are used. Such a viewpoint is a viewpoint derived for the purpose of suppressing the homologous recombination and increasing the production amount of the target protein. The first aspect relates to "codon match". Specifically, the minimum value of the codon matching index of CDS, which is a base sequence possessed by each individual and is a target of amino acid translation, is used as a reference. This is the first evaluation standard. And, the second aspect relates to the "base sequence of codon". Furthermore, the second aspect is divided into “number of unmatched bases” and “longest common string”. And it is a 2nd evaluation standard that it is based on the minimum of the number of unmatched bases among two CDS contained in individual data. Further, among the CDSs included in the individual data, the third evaluation standard is based on the length of the longest common character string. The significance of these three evaluation criteria will be described below.

(First evaluation criteria: degree of codon matching)
The first evaluation criterion, which is the first aspect, relates to "codon suitability". Here, “codon suitability” is higher as the number of frequently used codons is included in the CDS included in the individual data. Specifically, the minimum value (hereinafter referred to as minimum CAI value) of the codon adaptation index (hereinafter referred to as CAI) of CDS contained in each individual data is used as a reference. The CAI can be determined, for example, by the following equation.

L: Number of codons contained in individual data fi: Frequency of i-th codon max (fj): Frequency of synonymous codon (j-th codon) that is most frequent Here, synonymous codon means the same amino acid Codons that encode and have different sequences.

  Then, using the CAI determined by the above equation, the minimum CAI value can be determined by the following equation.

x: Number of CDS contained in individual data Ci: i-th CDS
CAI (Ci): Ci of CAI

  Here, it indicates that the higher the CAI of a certain CDS, the more the codons contained in the CDS (in other words, the fewer the number of rare codons contained in the CDS). And, if certain individual data has a CDS with an extremely low CAI (in other words, rich in rare codons), the CDS may not be translated efficiently. Therefore, by using the minimum CAI value as the first evaluation criterion and increasing the evaluation of such individual data as the minimum CAI value is larger, individual data having an extremely low CAI CDS can be removed in the optimization process. It will be possible. Therefore, more favorable simulation results can be obtained by increasing the minimum CAI value by the first mutation treatment.

(Second evaluation criteria: number of mismatched bases)
Next, the second evaluation criteria will be described with reference to FIG. The second evaluation criterion, which is the first of the second aspects, relates to the “number of unmatched bases”. Specifically, the minimum value of the number of unmatched bases (hereinafter referred to as the minimum number of unmatched bases) is used as an evaluation criterion. Here, the unmatched base refers to a base in which bases constituting codons do not match with each other between two CDS (hereinafter referred to as a CDS pair) Ci and Cj among x CDS included in individual data. It is. In the example of FIG. 13, the number of unmatched bases is five among the bases constituting Ci and Cj. Therefore, the number of unmatched bases of such a CDS pair (Ci and Cj) is five. The minimum number of unmatched bases can be determined by the following equation.

x: Number of CDS contained in individual data Ci: i-th CDS
Cj: jth CDS
NN (Ci, Cj): Number of unmatched bases of Ci and Cj

  Here, if one individual data has a CDS pair with an extremely low number of mismatched bases (in other words, similar base sequences), the possibility of homologous recombination between the CDS pairs increases. This is because the homologous recombination occurs at a site where the base sequences closely resemble (homologous site). Therefore, by using the minimum number of unmatched bases as the second evaluation criterion, and the higher the minimum number of unmatched bases (in other words, the larger the percentage of different base sequences), the base sequence is similar by raising the evaluation of such individual data. Individual data can be removed in the process of optimization. Therefore, more favorable simulation results can be obtained by increasing the number of unmatched bases by the second mutation treatment.

(Third evaluation criteria: longest common string)
Next, the third evaluation criteria will be described with reference to FIG. The second evaluation criterion, which is the second of the second aspects, relates to the "longest common character string". As described above, “longest common character string” refers to a character string representing codons contained at different sites within each CDS or one CDS, in contrast to the character strings contained in other CDSs. The longest matching string.

  Here, when the “exactly identical nucleotide sequence” is in the vicinity of the genome, the possibility of occurrence of homologous recombination is high. This is because, as described above, homologous recombination occurs at a site where the base sequences closely resemble (homologous site). Therefore, by using the length of the “longest common character string” as the third evaluation criterion and making the evaluation of such individual data higher as the length of the longest common character string is shorter, the proportion of “exactly the same base sequence” is high It is possible to remove individual data included in the process of optimization. Therefore, more favorable simulation results can be obtained by reducing the longest common character string by the second mutation process.

  As described above, it is possible to select individual data having a CDS in which a large number of frequently used codons are included by using the first evaluation criterion which is the first aspect. Further, by using the second evaluation criterion and the third evaluation criterion, which are the second aspect, it is possible to select individual data having a large proportion of different base sequences, and to suppress the occurrence of homologous recombination.

  Returning to FIG. 8, the main routine will be further described. In S26, the individual selecting unit 107 selects p individual data in descending order of rank from the g-th generation integrated data (including 2p individual data) classified for each front (per rank) in the Pareto optimal solution Do. Then, the parent individual group data acquisition unit 102 newly sets the selected p individual data as the (g + 1) th generation parent individual group data, and generates the (g + 1) th generation parent individual group data.

  Next, in S27, the processing unit 10 determines whether or not the variable g exceeds a predetermined number of generations G. And if this determination result is NO, it will progress to S28. On the other hand, if the determination result in S27 is YES, the main routine is ended, and the calculation result is output to the calculation data storage unit 204.

  If the determination result in S27 is NO, the process proceeds to S28, the variable g is incremented (that is, 1 is added to the variable g), and the process returns to S21 again. Here, since the variable g = 2 at the present time, the parent individual group data acquisition unit 102 acquires the second generation parent individual group data generated in S26. This process is repeated until the variable g reaches a predetermined number G of generations. In other words, the processing in S21 to S26 is repeated 250 times.

  By repeatedly executing the main routine described above, the p individual data items selected based on the three evaluation criteria become preferable as a gene sequence group as the number of repetitions increases.

<Example>
Examples of gene sequence design using the present algorithm will be described below. In this example, 10 CDSs encoding human insulin A chain (amino acid sequence: GIVEQCCTSICSLYQLENYCN) were designed as a simulation. The various parameters are as follows.
Predetermined probability Pm (mutation rate) = 0.05
Pc (crossing rate) = 0.5
Number of individual data included in individual population data of g-group (parent individual population data, child individual population data) p = 100
Number of predetermined generations G (maximum number of generations) = 250

  Hereinafter, with reference to FIG. 15 and FIG. 16, regarding the calculation results in this simulation, a graph plotting the first evaluation standard on the horizontal axis and the second evaluation standard on the vertical axis, and the first evaluation standard on the horizontal axis A graph in which the evaluation criteria are plotted on the vertical axis will be described.

  FIG. 15 is a graph in which the first evaluation criterion is plotted on the horizontal axis and the second evaluation criterion is plotted on the vertical axis. Here, in the graph, a plot represented by a circle indicates a first generation, a plot represented by a square indicates a tenth generation, and a plot represented by a triangle indicates a calculation result in the 250th generation. One plot corresponds to one design result (= solid data). As described above, since the first evaluation criterion is evaluated higher as the minimum CAI value is larger, the points plotted on the right side of the horizontal axis in the graph are better evaluated, and the points plotted on the left side of the horizontal axis are evaluated It can be said that it is bad. In the second evaluation criterion, the higher the minimum number of mismatched bases, the higher the evaluation. Therefore, in the graph, the higher the point plotted on the vertical axis, the better the evaluation, and the lower the vertical axis, the better. It can be said that it is bad. As shown in FIG. 15, as the number of generations increases (in other words, as the number of repetitions of the main routine in FIG. 8 increases), it can be read that the whole population data is preferable.

  FIG. 16 is a graph in which the first evaluation criterion is plotted on the horizontal axis and the third evaluation criterion is plotted on the vertical axis. The meanings of the plots represented by circles, squares and triangles are the same as in FIG. Here, in the third evaluation criteria, since the evaluation is higher as the length of the longest common string is shorter, the points plotted on the lower side of the vertical axis in the graph are evaluated better and plotted on the upper side of the vertical axis It can be said that the point is bad evaluation. As shown in FIG. 16, as the number of generations increases (in other words, as the number of repetitions of the main routine in FIG. 8 increases), it can be read that the entire population data is preferable.

  Although the various embodiments have been described above, the present invention is not limited to these.

  For example, the selection in S26 of the main routine in FIG. 8 uses the first evaluation criterion and the second evaluation criterion, or any one of the first evaluation criterion and the third evaluation criterion, and the graphs shown in FIG. It is also possible to obtain one of them. In addition, when using both the first and second evaluation criteria and both the first and third evaluation criteria, high evaluation individual data is obtained from the graph in FIG. 15 and the graph in FIG. 16 for each generation. The individual data may be selected and given points on any basis, and the total sum of points in these two graphs may be selected. Alternatively, instead of a two-dimensional graph as shown in FIG. 15 and FIG. 16, a three-dimensional graph with the first evaluation criterion as x axis, the second evaluation criterion as y axis, and the third evaluation criterion as z axis By creating, individual data that has obtained high evaluation for each of the three evaluation criteria may be selected.

  In addition, the storage unit 20 can be in an aspect of cloud computing provided in an information processing apparatus such as an external PC or a server without being provided inside the information processing apparatus 1. In this case, the external information processing apparatus transmits data necessary for each calculation to the information processing apparatus 1.

  Also, it can be provided as an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a dynamic reconfigurable processor (DRP) in which the functions of the information processing apparatus 1 are implemented. Further, the program may be provided to a computer as a program for realizing the functions of the information processing apparatus 1. In this case, such a program can also be distributed via the Internet or the like.

  Furthermore, "NSGA-II" which is a multipurpose genetic algorithm can also be used as this algorithm. This is because p optimal solutions can be obtained collectively as in the present algorithm. In addition, "simulated annealing" or "(single purpose) genetic algorithm", which is a kind of combined optimization algorithm, may be used. However, in this case, since the p optimal solutions can not be obtained collectively, it is necessary to repeat the calculation at least p times or more to obtain p optimal solutions. Furthermore, calculation results of these two or more algorithms may be mixed. In this case, an arbitrary integer α equal to or less than p is set, α individuals are selected from calculation results by a certain algorithm, and p−α individuals are selected from calculation results by another algorithm, and these are combined p It is also possible to use an individual.

Furthermore, the present invention
Data generated based on data representing an amino acid sequence, gene number and codon frequency table, which is a first generation parent individual population data representing a first generation parent individual population including a predetermined number of individual data Parent individual group data acquisition step to be acquired;
Performing mutation processing on the individuals included in the first generation parent individual population data;
Child individual population data acquisition step for acquiring first generation child individual population data representing a first generation child individual population including individuals subjected to the mutation process;
First-generation integration integrating the first-generation parent individual population data and the first-generation child individual population data, which is a predetermined evaluation criterion, based on codon suitability and an evaluation criterion regarding the base sequence of the codon A non-dominant sort execution step of executing non-dominant sort processing on data and classifying all individual data included in the first generation integrated data according to rank in Pareto optimal solution;
An individual selection step of selecting a predetermined number of the individual data in descending order of the rank from all individual data classified according to the rank;
It can also be understood as a gene sequence design method having

  1: Information processing apparatus 10: Processing unit 20: Storage unit 30: Operation unit 40: Display unit 50: Communication unit 100: Bus 101: Individual generation unit 102: Parent individual group data acquisition unit 103: child individual population data acquisition unit, 104: cross processing unit, 105: mutation processing unit, 106: non-dominated sort execution unit, 107: individual selection unit, 201: amino acid sequence data storage unit, 202: gene number data storage unit , 203: codon frequency table data storage unit, 204: calculation data storage unit, 205: evaluation criteria storage unit 205

Claims (16)

Data generated based on data representing an amino acid sequence, gene number and codon frequency table, which is a first generation parent individual population data representing a first generation parent individual population including a predetermined number of individual data Parent individual group data acquisition unit to acquire;
A mutation processing unit that executes mutation processing on the individuals included in the first generation parent individual population data;
A child individual population data acquisition unit for acquiring first generation child individual population data representing a first generation child individual population including individuals subjected to the mutation process;
First-generation integration integrating the first-generation parent individual population data and the first-generation child individual population data, which is a predetermined evaluation criterion, based on codon suitability and an evaluation criterion regarding the base sequence of the codon A non-dominant sort execution unit that executes non-dominant sort processing on data and classifies all individual data included in the first generation integrated data according to rank in a Pareto optimal solution;
An individual selecting unit which selects a predetermined number of the individual data in descending order of the rank from all individual data classified according to the rank;
An information processing apparatus having When the individual selecting unit selects the predetermined number of the individual data, if there is the individual data having the same rank, the individual selecting unit sequentially selects in order from the one with the largest crowded distance.
An information processing apparatus according to claim 1. The parent individual population data acquisition unit uses the individual data selected by the individual selection unit as second generation parent individual population data representing a second generation parent individual population.
The processing by the mutation processing unit, the non-dominated sort execution unit, and the individual selection unit is performed until a predetermined number of generations is reached.
The information processing apparatus according to claim 1 or 2. The evaluation standard for the degree of codon suitability is a base sequence possessed by each individual, and is based on the minimum value of the codon suitability index of CDS representing the base sequence to be subjected to amino acid translation.
The information processing apparatus according to any one of claims 1 to 3. The higher the minimum value of the codon matching index included in the individual, the higher the evaluation of the individual,
The information processing apparatus according to claim 4. The evaluation criteria for the base sequences of the codons are based on the minimum value of the number of unmatched bases representing the number of unmatched bases among the two CDSs included in each individual.
The information processing apparatus according to any one of claims 1 to 5. The higher the minimum value of the number of unmatched bases, the higher the evaluation of the individual,
The information processing apparatus according to claim 6. The evaluation criteria for the nucleotide sequences of the codons are the longest among the CDSs contained in each individual, among the CDSs among the individual CDSs or within a single CDS, the longest base sequence among the base sequences which are consecutively matched. Based on the length of the common string,
The information processing apparatus according to any one of claims 1 to 7. The shorter the length of the longest common string, the higher the value of the individual.
The information processing apparatus according to claim 8. The mutation processing unit
A second mutation process different from the first mutation process and the first mutation process is performed on each individual data included in the g-th generation parent individual population data representing the g-th generation parent individual population.
The information processing apparatus according to any one of claims 1 to 9. The mutation processing unit
Performing a first mutation process of replacing the codons contained in the CDS with codons more frequently than the codons with respect to all the CDSs contained in the individual, with a predetermined probability;
The information processing apparatus according to claim 10. The mutation processing unit
Among the CDSs contained in each individual, the codons overlapping with the longest common character, which is the longest base sequence among the base sequences which continuously coincide at different sites within each CDS or within one CDS, are predetermined. Execute a second mutation process that substitutes for another codon with a certain probability
The information processing apparatus according to claim 10. The first mutation treatment or the second mutation treatment is randomly selected.
The information processing apparatus according to any one of claims 10 to 12. And a crossover processing unit that performs crossover processing on the individuals included in the first generation parent individual population data,
The cross processing is
A predetermined even number of individual data is extracted from the g-th generation parent individual population data representing the g-th generation parent individual population, two individual data are selected from the extracted individual data, and the selected one is selected Perform cross processing on two sets of individual data,
The information processing apparatus according to any one of claims 1 to 13. The intersection processing unit
Determine a crossing point from boundaries of the codons included in the CDS included in the first individual data and the second individual data which are the two selected individual data,
The codons included in the first individual data and the second individual data are switched at the intersection point,
The information processing apparatus according to claim 14. Computer,
Data generated based on data representing an amino acid sequence, gene number and codon frequency table, which is a first generation parent individual population data representing a first generation parent individual population including a predetermined number of individual data Parent individual group data acquisition unit to acquire,
A mutation processing unit that executes mutation processing on the individuals included in the first generation parent individual population data,
A child individual population data acquisition unit for acquiring first generation child individual population data representing a first generation child individual population including individuals subjected to the mutation processing;
First-generation integration integrating the first-generation parent individual population data and the first-generation child individual population data, which is a predetermined evaluation criterion, based on codon suitability and an evaluation criterion regarding the base sequence of the codon A non-dominated sort execution unit that performs non-dominated sort processing on data and classifies all individual data included in the first generation integrated data according to rank in Pareto optimal solution;
An individual selection unit which selects a predetermined number of the individual data in descending order of the rank from all individual data classified according to the rank,
Information processing program to function as.