JP2006072861A - Transcription factor analysis program and transcription factor analysis method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To extract combination expected to have large relevance to expression of a gene from combination of a large number of transcription factors to perform efficient investigation. <P>SOLUTION: On coordinates wherein magnitude of a two-phenomenon mutual information amount showing coocurrence of a first transcription factor and a second transcription factor is taken as a horizontal axis and wherein magnitude of a three-phenomenon mutual information amount showing tissue peculiarity of the first transcription factor and the second transcription factor is taken as a vertical axis, the calculated mutual information amount is plotted and visualized to simultaneously analyze the coocurrence and the tissue peculiarity. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a transcription factor analysis program and a transcription factor analysis method for analyzing the relationship between gene expression and a transcription factor, and in particular, the efficiency of what is expected to be deeply related to gene expression from a combination of a large number of transcription factors. The present invention relates to a transcription factor analysis program and a transcription factor analysis method that can be well extracted.

  Living organisms have genetic information unique to each species. This genetic information exists on the genome, and all cells store the same information in the form of chromosomes. Recent research has revealed that the human genome, which is the human genome, consists of about 3 billion base pairs.

  Genes are scattered on the DNA constituting the chromosome and store control information related to protein production. In the case of humans, there are about 30,000 genes, but in reality, not all genes are expressed in a single cell to produce a protein. For example, in the cells of the heart, only genes that produce proteins necessary for the activity as the heart function. In this way, the gene is controlled so that only a part thereof functions for each living tissue.

  The mechanism for adjusting the function of genes for each living tissue has not yet been fully elucidated. At present, it has become the established theory that a substance called a transcription factor is involved, and it is actively researched which transcription factor is associated with which gene to produce which protein. A transcription factor is a substance that causes expression of a gene by binding to the upstream portion of the gene, and a certain number of transcription factors acting alone are already found.

  In a commercial product called TRANSFAC introduced in Non-Patent Document 1, it is possible to predict to some extent which gene upstream a known transcription factor binds. By using this product, it is possible to predict which transcription factor binds to the upstream part of a given gene.

“Mitsui Information Development Co., Ltd .: Product Introduction: Transcription Factors and Transcription Factor Binding Site Database“ TRANSFAC ”Essential for Gene Expression Research”, [online], August 26, 2004 [Search August 26, 2004] ] Internet <URL: http://bio.mki.co.jp/product/transfac/>

  However, research on the relationship between gene expression and transcription factors cannot be carried out efficiently simply by predicting which transcription factor binds to the upstream part of the gene.

  It is known that most genes are expressed not by a single transcription factor but by a combination of transcription factors. For this reason, once it is known which transcription factor binds to the upstream part of the gene, it is necessary to identify a combination that is actually related to expression from the combination of those transcription factors. . Since there are many combinations of transcription factors, this work requires a great deal of time and effort, and hinders the progress of research.

  The present invention has been made to solve the above-mentioned problems caused by the prior art, and extracts what is expected to be deeply related to gene expression from a combination of a large number of transcription factors to efficiently investigate. It is an object of the present invention to provide a transcription factor analysis program and a transcription factor analysis method that can be performed.

  In order to solve the above-described problems and achieve the object, a transcription factor analysis program according to the present invention is a transcription factor analysis program for analyzing the relationship between gene expression and transcription factor, and is a first in gene expression. A binding probability calculation procedure for calculating a binding probability between the transcription factor, the second transcription factor, and the biological tissue, a first transcription factor based on the binding probability calculated by the binding probability calculation procedure, A two-event mutual information calculation procedure means for calculating mutual information of two events representing the co-occurrence of transcription factors; a first transcription factor based on the binding probability calculated by the binding probability calculation procedure; 3 event mutual information calculation procedure means to calculate the mutual information of 3 events representing the tissue specificity of the transcription factor, and two dimensions taking the co-occurrence on one axis and the tissue specificity on the other axis 2 event phases on the coordinates of Generate a chart plotting the mutual information information calculated by the information amount calculation procedure means and the three-event mutual information amount calculation procedure means, and display an analysis result display processing procedure for displaying the chart on a predetermined display means. It is made to perform.

  The transcription factor analysis method according to the present invention is a transcription factor analysis method for analyzing the relationship between gene expression and transcription factor, wherein the first transcription factor, the second transcription factor in gene expression, A connection probability calculation step of calculating a connection probability with a biological tissue, a first transcription factor based on the connection probability calculated by the connection probability calculation step, and two events representing the co-occurrence of the second transcription factor 2 event mutual information calculation process means for calculating mutual information, 3 events representing the tissue specificity of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation process 3 event mutual information calculation process means for calculating mutual information of the two events, and the calculation of the two events mutual information on a two-dimensional coordinate taking the co-occurrence on one axis and the tissue specificity on the other axis Process means and 3 event mutual information calculator Generating a chart plotting the mutual information amount information calculated by means, characterized in that the chart had an analysis result display processing step of displaying on predetermined display means.

  According to this invention, the mutual information of two events representing the co-occurrence of a combination of transcription factors and the mutual information of three events representing the tissue specificity of the combination of transcription factors are simultaneously visualized on two-dimensional coordinates. Thus, a combination of transcription factors closely related to tissue-specific gene expression can be easily extracted.

  The transcription factor analysis program according to the present invention is the above-described invention, wherein the analysis result display processing procedure sets the mutual information amount if one of the mutual information amounts on the chart displayed on the display means is selected. Such detailed information is displayed on the display means.

  According to the present invention, since the detailed information such as the probability distribution by tissue of the combination of the selected transcription factors is displayed on the display means, the transcription factors extracted as deeply related to the tissue-specific gene expression are displayed. It is easy to analyze which tissue gene expression the combination is actually associated with.

  The transcription factor analysis program according to the present invention, in the above invention, further causes the computer to execute an unknown transcription factor registration procedure for receiving registration of an unknown transcription factor name and base sequence information. It is further characterized in that a binding probability obtained by applying the transcription factor registered by the unknown transcription factor registration procedure to one or both of the first transcription factor and the second transcription factor is further calculated.

  According to the present invention, since information on unknown transcription factors is registered and can be analyzed in the same manner as known transcription factors, whether or not the registered unknown transcription factors are related to tissue-specific gene expression. Thus, it is possible to predict whether or not the unknown transcription factor is actually a transcription factor.

  The transcription factor analysis program according to the present invention, in the above invention, further causes the computer to execute an unknown transcription factor generation procedure for generating base sequence information of an unknown transcription factor from genomic information, and the binding probability calculation procedure includes: The binding probability obtained by applying the transcription factor generated by the unknown transcription factor generation procedure to one or both of the first transcription factor and the second transcription factor is further calculated. The transcription factor analysis program described in 1. is characterized.

  According to the present invention, since information on unknown transcription factors is automatically generated and can be analyzed in the same manner as known transcription factors, the generated unknown transcription factors are related to tissue-specific gene expression. Whether or not the unknown transcription factor is actually a transcription factor can be predicted.

  According to the present invention, the mutual information of two events representing the co-occurrence of a combination of transcription factors on two-dimensional coordinates and the mutual information of three events representing a combination of tissue specificities of transcription factors are simultaneously visualized. Since it comprised as mentioned above, there exists an effect that the combination of a transcription factor deeply related with tissue-specific gene expression can be extracted easily.

  Further, according to the present invention, the detailed information such as the probability distribution by tissue of the combination of the selected transcription factors is displayed on the display means, so that the transcription extracted as deeply related to the tissue-specific gene expression The effect is that it is possible to easily analyze which tissue gene expression is actually associated with the combination of factors.

  In addition, according to the present invention, since information on unknown transcription factors is registered and can be analyzed in the same manner as known transcription factors, the registered unknown transcription factors are related to tissue-specific gene expression. Whether or not the unknown transcription factor is actually a transcription factor can be predicted.

  In addition, according to the present invention, information on unknown transcription factors is automatically generated and can be analyzed in the same manner as known transcription factors, so that the generated unknown transcription factors are related to tissue-specific gene expression. Whether or not the unknown transcription factor is actually a transcription factor can be predicted.

  Exemplary embodiments of a transcription factor analysis program and a transcription factor analysis method according to the present invention will be explained below in detail with reference to the accompanying drawings. Here, the case where the transcription factor analysis program and the transcription factor analysis method according to the present invention are used for the analysis of a transcription factor related to human gene expression will be described. The transcription factor analysis program and the transcription factor analysis method according to the present invention Can also be used to analyze transcription factors associated with gene expression in other organisms.

  First, the information theory on which the transcription factor analysis method according to this embodiment is based will be described. In the transcription factor analysis method according to the present embodiment, two types of mutual information, that is, a mutual information amount of two events and a mutual information amount of three events are evaluated.

  The mutual information amount of two events represents the co-occurrence between two factors and has been studied since the middle of the last century. For example, it is used to measure the co-occurrence of words in a sentence and is applied to the determination of the priority order of conversion candidates in a kanji conversion system.

  Specifically, the mutual information amount I (X; Y) of two events is obtained by the following formula 1.

  Here, H (X) is called Shannon's information amount and is expressed by the following Equation 2.

  On the other hand, the mutual information amount of the three events is calculated by Equation 3 below.

  Although it has been suggested that the mutual information of these three events can be used for analysis of so-called “complex systems”, it has been rarely used in practice. In the transcription factor analysis method according to the present example, attention is paid to the fact that the mutual information amount of the three events expresses the specificity of the relationship between the factors, and it was decided to use it simultaneously with the mutual information amount of the two events.

  FIG. 1 is a sample diagram showing an application example of the transcription factor analysis method according to the present embodiment. As shown in the figure, in the transcription factor analysis method according to this embodiment, the mutual information amount (co-occurrence) of two events is taken on one axis, and the mutual information amount (specificity) of three events is taken on the other axis. Plot the combinations of various transcription factors on the coordinates.

  The mutual information amount of two events represents the high level of co-occurrence of the combination of two types of transcription factors. This co-occurrence becomes larger as the two types of transcription factors frequently bind to the upstream portions of various genes simultaneously (or do not bind simultaneously).

  On the other hand, the mutual information amount of the three events represents the high specificity of the combination of two types of transcription factors and biological tissues. In general, the mutual information of the three events can take either positive or negative values. In this example, a negative value is shown. It is mathematically suggested that this value indicates various states of the relationship between the three factors. In any case, it is understood that the larger the absolute value, the greater the influence of the third factor.

  For example, if the mutual information value of three events is 0 and the co-occurrence of the first factor and the second factor is high, it indicates that there is almost no influence of the third factor. Conversely, when the value of the mutual information amount of the three events is large, the details of the influence are unknown, but the influence of the third factor is so large. In this embodiment, since a biological tissue event is adopted as the third factor, the mutual information amount of the three events serves as an index indicating the specificity of the biological tissue, and the magnitude of the absolute value is specific. It is understood that it shows the size.

The following are reference materials related to the mutual information of the three events.
[Reference 1] Tsujishita, T., "On Triple Mutual Information", ADVANCED IN APPLIED MATHEMATICS, 1995: p.269-274

  A combination of transcription factors having high co-occurrence and low specificity, such as the combination plotted in region 1 of FIG. 1, has a similar relationship between transcription factors regardless of the expression tissue for any gene. Indicates that it exists. It can be judged that the combination of transcription factors having such a general relationship is unlikely to be related to the expression of a specific gene, and it is understood that there is a high possibility that the influence due to the homology of the gene sequence is strongly expressed. . This is because the determination of gene binding between transcription factors is performed for each transcription factor, so if the sequence homology is high, the same binding determination is likely to be made, resulting in high co-occurrence and specificity. This is because it tends to be lower.

  Moreover, like the combination plotted in the region 2 of FIG. 1, a combination of transcription factors with high specificity is highly likely to have a specific relationship with a gene expressed in a specific living tissue. A combination of transcription factors having such a specific relationship is likely to be related to the expression of a specific gene, and it can be judged that it is appropriate to raise the priority as a research object. In particular, a factor with low co-occurrence and high specificity is expected to be very likely to be associated with the expression of a specific gene because it has no co-occurrence effect.

  In this way, by co-occurrence and specificity are evaluated at the same time, and by plotting the evaluation results on a two-dimensional coordinate space with both as axes, a specific biological tissue can be selected from a number of combinations of transcription factors. Combinations with specific causal relationships can be extracted, which is a useful index for improving research efficiency.

  Next, the configuration of the transcription factor analyzer according to this embodiment will be described. FIG. 2 is a functional block diagram showing the configuration of the transcription factor analyzer according to the present embodiment. As shown in the figure, the transcription factor analyzer 200 is connected to the DB server 100 via a network or the like.

  The DB server 100 is a DB server that stores various information necessary for analysis of transcription factors, and includes a transcription factor sequence information DB 110, a gene expression information DB 120, a genome sequence information DB 130, and a gene mapping information DB 140.

  The transcription factor sequence information DB 110 is a DB that stores base sequence information of known transcription factors. FIG. 3 is a sample diagram showing an example of the data configuration of the transcription factor sequence information DB 110 shown in FIG. 2, and shows base sequence information of the transcription factor XXX1.

The figure shows that the transcription factor XXX1 is a sequence composed of three bases, and the base appearance probability in each component is as follows.
First element: 100% cytosine (C)
Second element: guanine (G) 50%, adenine (A) 50%
Third element: guanine (G) 10%, adenine (A) 10%, thymine (T) 80%

  The reason why the occurrence probability of bases in the second and third elements is not constant is that there is fluctuation in the base sequence of the transcription factor.

  The gene expression information DB 120 is a DB that stores gene expression information for each living tissue measured by experiments or the like. FIG. 4 is a sample diagram showing an example of the data configuration of the gene expression information DB 120 shown in FIG.

  In this example, the type of gene is identified by HIT which is the ID of cDNA, but it may be identified using a genetic locus. In this example, the expression information is expressed as a ratio that is a relative value, but may be expressed as an absolute value.

  The genome sequence information DB 130 is a DB that stores base sequence information of a genome (human genome in this embodiment). The gene mapping information DB 140 is a DB that stores which gene is present at which position on the genome.

  In FIG. 2, all DBs are illustrated as being stored in a single DB server, but a configuration in which DBs are distributed among a plurality of DB servers may be employed. Further, the transcription factor analyzer 200 may be configured to incorporate any or all of the DBs.

  The transcription factor analysis device 200 is a device that analyzes the relationship between a combination of transcription factors and a living tissue according to conditions specified by the user, and presents the analysis result to the user. The input factor 210, the display unit 220, and an interface Unit 230, control unit 240, and storage unit 250.

  The input unit 210 is a device that accepts user input and includes a keyboard and a mouse. The display unit 220 is a device that displays image data, character data, and the like on a screen, and a liquid crystal display device or the like corresponds to this. The interface unit 230 is an interface for exchanging data with the DB server 100 through a network or the like.

  The control unit 240 is a control unit that controls the transcription factor analysis apparatus 200 as a whole, and includes a condition input receiving unit 240a, a combination determination processing unit 240b, a discretization processing unit 240c, a probability processing unit 240d, and a joint probability calculation. Unit 240e, two-event mutual information amount calculation unit 240f, three-event mutual information amount calculation unit 240g, analysis result display processing unit 240h, and information acquisition unit 240i.

  The condition input receiving unit 240a is a processing unit that displays a condition input screen on the display unit 220 and receives input of various conditions for analysis from the user. In the transcription factor analysis according to the present embodiment, one of the combinations of transcription factors is fixed and the analysis is performed, and the condition input receiving unit 240a determines the type of transcription factor to be fixed and analyzed as a part of the condition. Accept instructions.

  The binding determination processing unit 240b determines whether or not the transcription factor binds to the upstream portion of various genes stored in the gene mapping information DB 140, and stores the determination result in the binding determination result storage unit 250a of the storage unit 250. It is a processing unit.

  FIG. 5 is a sample diagram illustrating an example of a data configuration of the combination determination result storage unit 250a illustrated in FIG. As shown in the figure, if the binding determination processing unit 240b determines that the designated transcription factor binds to the upstream portion of any gene, it stores the transcription factor and gene information here.

  The discretization processing unit 240c is a processing unit that discretizes the expression information in the gene expression information DB 120 to create a class value. Expression information is generally expressed as a relative value as a ratio to a standard marker, and is used to compare the expression level of the same gene for each living tissue. For this reason, it is generally difficult to compare the level of expression between different genes. Therefore, the discretization processing unit 240c discretizes gene expression information, enables comparison of expression levels between different genes, and stores the result in the discretization result storage unit 250b of the storage unit 250.

  FIG. 6 is a sample diagram illustrating an example of a data configuration of the discretization result storage unit 250b illustrated in FIG. In this example, the expression information shown in FIG. 4 is binarized with 30% as a threshold, more than 1 and less than 0. The threshold value can be set to an arbitrary size. Also, it is possible to discretize to more than three values by providing a plurality of threshold values.

  The probability processing unit 240d combines the processing result of the combination determination processing unit 240b and the discretization processing unit 240c, calculates the probability amount of each factor, and stores it in the probability result storage unit 250c of the storage unit 250. It is. Specifically, a process of calculating the probability by classifying and totaling the number of gene expressions by biological tissue created by the discretization processing unit 240c by discretization for each binding pattern found by the binding determination processing unit 240b. Do it.

  FIG. 7 is a sample diagram illustrating an example of a data configuration of the probability result storage unit 250c illustrated in FIG. Here, a portion storing the transcription factor XXX1 designated to be fixedly analyzed and the probability amount of YYY1, which is one of the transcription factors analyzed in combination, is extracted.

The connection probability calculation unit 240e is a processing unit that calculates a connection probability necessary for calculating the mutual information amount and stores it in the connection probability storage unit 250d of the storage unit 250. Specifically, it is a general expression of conditional probability.
P (X, Y, Z) = P (Z | X, Y) × P (X, Y)
From this, P (X, Y, Z) is obtained, and P (Y, Z) and P (Z, X) are calculated from this P (X, Y, Z).

  FIG. 8 is a sample diagram showing an example of the data configuration of the connection probability storage unit 250d shown in FIG. Here, extract the transcription factor XXX1 specified to be fixed and analyzed, YYY1 which is one of the transcription factors analyzed in combination with this, and the part storing the binding probability of tissues A to Z is doing.

  The two-event mutual information amount calculation unit 240f calculates the mutual information amount of two events of two transcription factors based on the information calculated by the binding probability calculation unit 240e, and stores the mutual information amount in the mutual information amount storage unit 250e of the storage unit 250. It is a processing part to be made. Further, the three-event mutual information calculation unit 240g calculates the mutual information amount of the three events of the two transcription factors and the living tissue based on the information calculated by the coupling probability calculation unit 240e, and the mutual information amount of the storage unit 250 This is a processing unit to be stored in the storage unit 250e.

  In order to calculate the mutual information amount, Formulas 1 to 3 are used as described above. In order to obtain the mutual information amount, it is necessary to obtain the Shannon information amount using Equation 2. Hereinafter, the mutual information amount will be calculated using the connection probabilities shown in FIG. 8 as X: XXX1, Y: YYY1, and Z: organization.

Since X takes {0, 1} two values, H (X) is
H (X) = − {0.5 × Log (0.5) + 0.5 × Log (0.5)}
=-{-0.3465657359-0.3465657359}
= 0.69314718
Is required.

  H (Y) is obtained in the same manner as H (X) and becomes 0.69314718. Since Z takes four values, {Organization A, Organization B, Organization C, Organization D}, H (Z) is 1.27038049 by the same calculation.

  (X, Y) takes four states {(1, 1), (1, 0), (0, 1), (0, 0)}, and H (X, Y) is H (X). It becomes 1.386294361 by the same calculation. (Y, Z) takes eight states combining Y = {0, 1} and Z = {A, B, C, D}, and H (Y, Z) is 1.900225615. H (Z, X) becomes 1.900225615 as H (Y, Z).

  Further, (X, Y, Z) has the 16 states shown in FIG. 8, and when H (X, Y, Z) is calculated, it becomes 2.31957997.

By applying the Shannon information amount to Equations 1 and 3, the mutual information amount of two events and the mutual information amount of three events are obtained as follows.
I (X; Y) = 0
I (X; Y; Z) = − 0.205831844

  FIG. 9 is a sample diagram illustrating an example of a data configuration of the mutual information storage unit 250e illustrated in FIG. As shown in the figure, the mutual information storage unit 250e includes a fixed transcription factor type, a transcription factor type combined with the fixed transcription factor, and two events obtained from the relationship between these transcription factors. The information amount and the mutual information amount of three events are stored.

  The analysis result display processing unit 240h is a processing unit that displays the calculated mutual information and various information obtained in the process of calculating the mutual information on the display unit 220 as analysis results. The screen displayed by the analysis result display processing unit 240h will be described later.

  The information acquisition unit 240 i is a processing unit that acquires various types of information from the DB server 100 via the interface unit 230.

  The storage unit 250 includes a combination determination result storage unit 250a, a discretization result storage unit 250b, a probability result storage unit 250c, a connection probability storage unit 250d, and a mutual information storage unit 250e. Since these storage units have already been described, description thereof is omitted here.

  Next, screens displayed on the display unit 220 by the condition input receiving unit 240a and the analysis result display processing unit 240h will be described. FIG. 10 is a sample diagram showing an example of the condition input screen. This screen is a screen for the condition input receiving unit 240a to receive input of various conditions for analysis from the user.

  In the "Specify transcription factor to be fixed" item at the bottom, select one transcription factor. In the transcription factor analysis method according to the present embodiment, analysis is performed with one of the combinations of transcription factors fixed, and the transcription factor selected here becomes the transcription factor to be fixed and analyzed.

  FIG. 11 is a sample diagram showing an example of the analysis result display screen. This screen is a screen on which the analysis result display processing unit 240h visualizes and displays the mutual information calculated by the two-event mutual information calculation unit 240f and the three-event mutual information calculation unit 240g.

  As shown in the figure, a graph data display area occupying most of the screen exists at the upper left of the screen. In this area, the co-occurrence (mutual information amount of two events) is taken on the horizontal axis, and the mutual information storage unit 250e is on the coordinates where the tissue specificity (mutual information amount of three events) is taken on the vertical axis. The mutual information stored in is plotted.

  When the user selects one of the plotted points, the detailed information of the transcription factor is displayed in the area to the right of the graph data display area. In the area below the graph data display area, there is provided an area in which various conditions designated for obtaining the mutual information are displayed.

  When the “distribution graph display” button is pressed on the analysis result display screen, the distribution graph display screen is displayed. FIG. 12 is a sample diagram illustrating an example of a distribution graph display screen. As shown in the figure, on this screen, the probability distribution for each tissue of the transcription factor selected on the analysis result display screen and the fixed transcription factor is displayed as a graph.

  On the screen, the expected value when the two transcription factors are uncorrelated is displayed as a dotted line. By comparing this dotted line with the graph, in which tissue there is a specific relationship between the two transcription factors Can be easily identified. Further, on this screen, by pressing the “discretization number” button, a graph of the discretization number can be displayed instead of the graph of the probability distribution.

  When a “gene list display” button is pressed on the analysis result display screen, a gene list display screen is displayed. FIG. 13 is a sample diagram showing an example of a gene list display screen. As shown in the figure, on this screen, the number of discretizations for each organization of the transcription factor selected on the analysis result display screen and the fixed transcription factor is displayed as a list.

  When the “display list data” button is pressed on the analysis result display screen, the list data display screen is displayed. FIG. 14 is a sample diagram showing an example of a list data display screen. As shown in the figure, on this screen, the mutual information amount and the discretization number of all transcription factor combinations are displayed.

  Next, the processing procedure of the transcription factor analyzer 200 shown in FIG. 2 will be described. FIG. 15 is a flowchart showing a processing procedure of the transcription factor analysis apparatus 200 shown in FIG.

  As shown in the figure, when input of various conditions for analysis is accepted (step S101), the binding determination of all genes and transcription factors is performed (step S102), and the expression data of the genes is discretized. (Step S103).

  Subsequently, one transcription factor to be combined with the designated transcription factor is selected (step S104). If all the transcription factors registered in the transcription factor sequence information DB 110 have been selected (Yes at Step S105), the process proceeds to Step S110.

  If all the transcription factors registered in the transcription factor sequence information DB 110 have not been selected (No at Step S105), the fixed transcription factor and the selected transcription factor are stochasticized if fixed, and the state of the transcription factor is determined. A probability distribution for each organization / organization is obtained (step S106), and a connection probability is calculated based on the probability distribution (step S107).

  If the connection probability is calculated, the mutual information amount of two events is calculated based on this (step S108), and the mutual information amount of three events is further calculated (step 109). In this way, when the mutual information amount between the designated transcription factor and the selected transcription factor is determined to be fixed, the processing is restarted from step S104.

  If all the transcription factors registered in the transcription factor sequence information DB 110 have been selected in step S104 (Yes in step S105), all the obtained mutual information amounts are plotted and visualized on the analysis result screen. This screen is presented to the user (step S110).

  As described above, in Example 1, the co-occurrence of transcription factor combinations is taken as one axis, and the mutual information is plotted and visualized on a coordinate space where the tissue specificity is taken as the other axis. Because it is configured to simultaneously analyze the co-occurrence and tissue specificity of transcription factor combinations, transcription factor combinations that are predicted to be closely related to gene expression in specific tissues can be easily identified, The efficiency of transcription factor research can be improved.

  In Example 1, although the example which analyzes the relationship between the combination of a known transcription factor and gene expression was demonstrated, the transcription factor analysis program and transcription factor analysis method concerning this invention are also in order to predict an unknown transcription factor. Available. In Example 2, a case where the transcription factor analysis program and the transcription factor analysis method according to the present invention are used for predicting an unknown transcription factor will be described.

  Not all transcription factors associated with gene expression are known, and there are unknown ones. If the transcription factor analysis program and the transcription factor analysis method according to the present invention are used, a combination of unknown transcription factors, or a combination of an unknown transcription factor and a known transcription factor is a gene expression and specificity in a specific living tissue. If an unknown transcription factor is found, it is likely that the unknown transcription factor is actually a transcription factor.

  Thus, if an unknown transcription factor can be easily predicted, the efficiency of transcription factor research can be greatly improved.

  The transcription factor analysis method described in Example 1 can cope with prediction of an unknown transcription factor without making a major change. In order to determine the binding of the transcription factor to the upstream part of the gene, it is sufficient if the base sequence information of the transcription factor is sufficient, and there is no problem even if the base sequence information of the transcription factor is unknown. In other treatments, there is no factor that the transcription factor must be known.

  Therefore, in the second embodiment, only differences from the first embodiment will be described. FIG. 16 is a functional block diagram showing the configuration of the transcription factor analyzer according to the present embodiment.

  As shown in the figure, the control unit 240 further includes an unknown transcription factor registration unit 240j. The unknown transcription factor registration unit 240j displays an unknown transcription factor registration screen on the display unit 220, causes the user to input the name and base sequence information of the unknown transcription factor, and inputs the input information to the unknown transcription factor in the storage unit 250. The processing unit is stored in the storage unit f.

  FIG. 17 is a sample diagram showing an example of the unknown transcription factor registration screen. As shown in the figure, it is possible to register the name of an unknown transcription factor and the appearance probability of a base in each base sequence.

  The storage unit 250 further includes an unknown transcription factor storage unit f. The data configuration of the unknown transcription factor storage unit f is the same as the data configuration of the transcription factor sequence information DB 110 described in FIG.

  In this embodiment, the user registers the base sequence of an unknown transcription factor. However, the program may automatically generate the base sequence of an unknown transcription factor. From the genome information, the base sequence of the upstream part of the gene has been clarified, and the base sequence of the transcription factor is an inversion of this part.

  Next, the processing procedure of the transcription factor analyzer 201 shown in FIG. 16 will be described. FIG. 18 is a flowchart showing a processing procedure of the transcription factor analyzer 201 shown in FIG.

  Here, only differences from the first embodiment will be described. Before accepting input of various analysis conditions, registration of an unknown transcription factor is accepted (step S201). When receiving input of various conditions, the transcription factor to be fixed can be selected not only from the known transcription factor but also from the transcription factor input in step S201. Also in the subsequent processing, the transcription factor input in step S201 is handled in the same manner as the known transcription factor.

  As described above, in Example 2, the base sequence information of an unknown transcription factor is registered (or generated), and the relationship between the unknown transcription factor and gene expression can be analyzed. Thus, transcription factors that actually function as transcription factors can be efficiently selected from unknown transcription factors.

  In Example 1 and Example 2, the example in which the method of simultaneously visualizing and analyzing the mutual information amount of two events and the mutual information amount of three events was used for the analysis of transcription factors was described. The use of the analysis method is not limited to this, and can be used for various analyses. This analysis method is particularly useful when the purpose is to find a unique relationship among events in which a plurality of factors are complicatedly related.

(Appendix 1) A transcription factor analysis program for analyzing the relationship between gene expression and transcription factors,
A binding probability calculation procedure for calculating the binding probability of the first transcription factor, the second transcription factor, and the biological tissue in the expression of the gene;
Two-event mutual information calculation procedure means for calculating the mutual information of two events representing the co-occurrence of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation procedure; ,
A three-event mutual information calculation procedure means for calculating a mutual information amount of three events representing the tissue specificity of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation procedure; ,
Mutually calculated by the two-event mutual information calculation procedure means and the three-event mutual information calculation procedure means on a two-dimensional coordinate taking the co-occurrence as one axis and the tissue specificity as the other axis. A transcription factor analysis program characterized by causing a computer to generate a chart on which information amount information is plotted and to display an analysis result display processing procedure for displaying the chart on a predetermined display means.

(Appendix 2) The analysis result display processing procedure is characterized in that if one of the mutual information amounts on the chart displayed on the display means is selected, the detailed information on the mutual information amount is displayed on the display means. The transcription factor analysis program according to appendix 1.

(Supplementary Note 3) The computer further executes an unknown transcription factor registration procedure for accepting registration of an unknown transcription factor name and nucleotide sequence information,
The binding probability calculation procedure further calculates a binding probability in which the transcription factor registered by the unknown transcription factor registration procedure is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis program according to appendix 1 or 2.
(Appendix 4) The computer further executes an unknown transcription factor generation procedure for generating base sequence information of an unknown transcription factor from genomic information,
The binding probability calculation procedure further calculates a binding probability in which the transcription factor generated by the unknown transcription factor generation procedure is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis program according to appendix 1 or 2.

(Appendix 5) A transcription factor analysis method for analyzing the relationship between gene expression and transcription factors,
A binding probability calculation step of calculating a binding probability of the first transcription factor, the second transcription factor, and the biological tissue in gene expression;
Two-event mutual information calculation step means for calculating a mutual information amount of two events representing the co-occurrence of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation step; ,
3-event mutual information calculation step means for calculating the mutual information of the first transcription factor and the three events representing the tissue specificity of the second transcription factor based on the binding probability calculated by the binding probability calculation step; ,
Mutually calculated by the two-event mutual information calculation process means and the three-event mutual information calculation process means on a two-dimensional coordinate taking the co-occurrence on one axis and the tissue specificity on the other axis. A transcription factor analysis method comprising: generating a chart plotting information amount information and displaying the chart on a predetermined display means.

(Appendix 6) The analysis result display processing step is characterized in that, when one of the mutual information amounts on the chart displayed on the display means is selected, detailed information relating to the mutual information amount is displayed on the display means. The transcription factor analysis method according to appendix 5.

(Additional remark 7) It further has the unknown transcription factor registration process which receives registration of the name and base sequence information of an unknown transcription factor,
The binding probability calculating step further calculates a binding probability obtained by applying the transcription factor registered in the unknown transcription factor registration step to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis method according to appendix 5 or 6.
(Additional remark 8) It further has the unknown transcription factor production | generation process which produces | generates the base sequence information of an unknown transcription factor from genome information,
The binding probability calculating step further calculates a binding probability obtained by applying the transcription factor generated by the unknown transcription factor generating step to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis method according to appendix 5 or 6.

(Appendix 9) A transcription factor analyzer for analyzing the relationship between gene expression and transcription factors,
A binding probability calculating means for calculating a binding probability of the first transcription factor, the second transcription factor, and the biological tissue in gene expression;
Two-event mutual information calculation means for calculating the mutual information of two events representing the co-occurrence of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation means; ,
Three-event mutual information calculation means for calculating mutual information of three events representing the tissue specificity of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation means; ,
The mutual events calculated by the two-event mutual information calculation means and the three-event mutual information calculation means on the two-dimensional coordinates taking the co-occurrence on one axis and the tissue specificity on the other axis. A transcription factor analyzing apparatus comprising: an analysis result display processing means for generating a chart in which information amount information is plotted and displaying the chart on a predetermined display means.

(Supplementary Note 10) The analysis result display processing means, when one of the mutual information amounts on the chart displayed on the display means is selected, displays detailed information on the mutual information amount on the display means. The transcription factor analyzer according to appendix 9.

(Supplementary note 11) An unknown transcription factor registration means for accepting registration of the name and base sequence information of an unknown transcription factor,
The binding probability calculation means further calculates a binding probability in which the transcription factor registered by the unknown transcription factor registration means is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analyzer according to appendix 9 or 10.
(Additional remark 12) The transcription | transfer element production | generation means which produces | generates the base sequence information of an unknown transcription factor from genome information is further provided,
The binding probability calculating means further calculates a binding probability in which the transcription factor generated by the unknown transcription factor generating means is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analyzer according to appendix 9 or 10.

  As described above, the transcription factor analysis program and the transcription factor analysis method according to the present invention are useful for the analysis of transcription factors, and in particular, those that are expected to be closely related to gene expression from combinations of a large number of transcription factors. It is suitable for the case where it is necessary to extract and efficiently investigate.

It is a sample figure which shows the example of application of the transcription factor analysis system concerning a present Example. It is a functional block diagram which shows the structure of the transcription factor analyzer based on a present Example. It is a sample figure which shows an example of a data structure of transcription factor arrangement | sequence information DB shown in FIG. It is a sample figure which shows an example of a data structure of gene expression information DB shown in FIG. It is a sample figure which shows an example of a data structure of the combination determination result memory | storage part shown in FIG. It is a sample figure which shows an example of a data structure of the discretization result memory | storage part shown in FIG. It is a sample figure which shows an example of a data structure of the probability result memory | storage part shown in FIG. It is a sample figure which shows an example of a data structure of the joint probability memory | storage part shown in FIG. It is a sample figure which shows an example of a data structure of the mutual information storage part shown in FIG. It is a sample figure which shows an example of a condition input screen. It is a sample figure which shows an example of an analysis result display screen. It is a sample figure which shows an example of a distribution graph display screen. It is a sample figure which shows an example of a gene list display screen. It is a sample figure which shows an example of a list data display screen. It is a flowchart which shows the process sequence of the transcription factor analyzer shown in FIG. It is a functional block diagram which shows the structure of the transcription factor analyzer based on a present Example. It is a sample figure which shows an example of an unknown transcription factor registration screen. It is a flowchart which shows the process sequence of the transcription factor analyzer shown in FIG.

Explanation of symbols

100 DB server 110 Transcription factor sequence information DB
120 Gene expression information DB
130 Genome sequence information DB
140 Gene mapping information DB
200 transcription factor analyzer 201 transcription factor analyzer 210 input unit 220 display unit 230 interface unit 240 control unit 240a condition input reception unit 240b coupling determination processing unit 240c discretization processing unit 240d randomization processing unit 240e coupling probability calculation unit 240f 2 events Mutual information calculation unit 240g 3-event mutual information calculation unit 240h Analysis result display processing unit 240i Information acquisition unit 240j Unknown transcription factor registration unit 250 Storage unit 250a Binding determination result storage unit 250b Discretization result storage unit 250c Stochasticization result storage unit 250d Binding probability storage unit 250e Mutual information storage unit 250f Unknown transcription factor storage unit

Claims (5)

A transcription factor analysis program for analyzing the relationship between gene expression and transcription factors,
A binding probability calculation procedure for calculating the binding probability of the first transcription factor, the second transcription factor, and the biological tissue in the expression of the gene;
Two-event mutual information calculation procedure means for calculating the mutual information of two events representing the co-occurrence of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation procedure; ,
A three-event mutual information calculation procedure means for calculating a mutual information amount of three events representing the tissue specificity of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation procedure; ,
Mutually calculated by the two-event mutual information calculation procedure means and the three-event mutual information calculation procedure means on a two-dimensional coordinate taking the co-occurrence as one axis and the tissue specificity as the other axis. A transcription factor analysis program characterized by causing a computer to generate a chart on which information amount information is plotted and to display an analysis result display processing procedure for displaying the chart on a predetermined display means.   The analysis result display processing procedure is characterized in that, when one of mutual information amounts on the chart displayed on the display means is selected, detailed information relating to the mutual information amount is displayed on the display means. The transcription factor analysis program described in 1. The computer further executes an unknown transcription factor registration procedure for accepting registration of an unknown transcription factor name and nucleotide sequence information,
The binding probability calculation procedure further calculates a binding probability in which the transcription factor registered by the unknown transcription factor registration procedure is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis program according to claim 1 or 2. The computer further executes an unknown transcription factor generation procedure for generating base sequence information of an unknown transcription factor from genomic information,
The binding probability calculation procedure further calculates a binding probability in which the transcription factor generated by the unknown transcription factor generation procedure is applied to one or both of the first transcription factor and the second transcription factor. The transcription factor analysis program according to claim 1 or 2. A transcription factor analysis method for analyzing the relationship between gene expression and transcription factors,
A binding probability calculation step of calculating a binding probability of the first transcription factor, the second transcription factor, and the biological tissue in gene expression;
Two-event mutual information calculation step means for calculating a mutual information amount of two events representing the co-occurrence of the first transcription factor and the second transcription factor based on the binding probability calculated by the binding probability calculation step; ,
3-event mutual information calculation step means for calculating the mutual information of the first transcription factor and the three events representing the tissue specificity of the second transcription factor based on the binding probability calculated by the binding probability calculation step; ,
Mutually calculated by the two-event mutual information calculation process means and the three-event mutual information calculation process means on a two-dimensional coordinate taking the co-occurrence on one axis and the tissue specificity on the other axis. A transcription factor analysis method comprising: generating a chart plotting information amount information and displaying the chart on a predetermined display means.

Images