JP2002107366A - Diagnosis-assisting system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system for assisting diagnosis of physicians by estimating the disease with which a patient is suffering currently or will probably suffer in the future, using information obtained from a microarray. SOLUTION: With reference to a microarray emission pattern database 11 and a disease relationship database 12, a disease-inferring means 14 infers the disease of high liable possibility which is then delivered to an I/O means 16. A disease-related gene searching means 15 searches for the relation between a microarray emission pattern and the disease with reference to the microarray emission pattern data base 11 and a gene knowledge database 13 and updates the content of the disease relationship database 12. Information for assisting diagnosis of intricate disease, having a variety of causes or a disease exhibiting difference morbid states from person to person, can be provided. Furthermore, diagnosis can be effected, based on up-to-date knowledge at all times by reflecting the results thus obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a diagnosis support system for assisting a diagnosis by a physician, and more particularly to a method for accurately estimating a disease which a patient currently has or a disease which is likely to be in the future by using a DNA microarray image. The present invention relates to a diagnosis support system that provides information for assisting a diagnosis by a doctor.

[0002]

2. Description of the Related Art DNA microarrays (hereinafter referred to as "microarrays") have a high density on a glass or silicon substrate.
NA is bound, the DNA is hybridized with a labeled probe prepared from RNA extracted from a tissue or cell, a signal is detected, and the intensity of this signal is measured to analyze gene expression in the tissue or cell. Things. With the development of the microarray, expression information of thousands to tens of thousands of genes can be comprehensively analyzed at once, and simultaneous measurement of gene expression information in all regions of the genome has become possible. In addition, with the breakthrough in genetic research, new genes have been discovered one after another in all regions of the genome, and by analyzing their functions, genes related to disease onset, severity, drug sensitivity, etc. are being identified for various diseases. is there.

[0003] On the other hand, in the field of medical treatment, the importance of diagnosing complicated diseases having a wide variety of causes and diseases having different disease states according to individuals is increasing. Since the onset and severity of such diseases are greatly related to the individual's constitution and living environment, genetic diagnosis to examine the expression level of disease-related genes and the presence or absence of mutations in individual patients is one of the effective means. Has become. Conventional gene diagnosis has been used to diagnose and predict the onset of disease by examining the expression level and the presence or absence of mutations in a small number of well-known genes, but the functions of human genes have been elucidated. This is only about 10% of the total, and in order to realize a diagnosis of a complex disease with a variety of causes or a disease whose condition varies from individual to individual, a large amount of genes, including genes whose association with the disease is unknown, is required. Techniques for comprehensively judging gene expression information have become indispensable.

[0004]

As described above, as described above,
Once the relationship between the large amount of gene expression information obtained from the microarray and the onset, severity, drug sensitivity, etc. of the disease is clarified, a database of those relationships will be used to make higher levels of information based on the information obtained from the microarray. It is possible to diagnose and predict disease onset with high accuracy,
It becomes possible to determine a treatment policy and prescribe a drug according to the individual patient. However, there has been a problem that a system for diagnosing or predicting the onset of disease based on information obtained from a microarray has not been established yet.

[0005]

In order to solve the above-mentioned problems, a diagnostic support system according to the present invention comprises a microarray light-emitting pattern database for storing a microarray light-emitting pattern, and a disease for storing a relationship between the microarray light-emitting pattern and a disease. A relational database, a gene knowledge database for storing knowledge about genes and diseases, input / output means for inputting and outputting results of microarray images, and image analysis for reading microarray images input from the input / output means and performing image analysis Means, and a disease inference means, wherein the disease inference means refers to the microarray luminescence pattern database and the disease association database, and from the microarray luminescence pattern, the disease that the patient is currently suffering from, or is likely to suffer in the future Infer disease and enter inference results Output via the force means. In addition, the patient name,
Basic patient data such as address, date of birth, family composition, patient history, family history, chief complaint, findings, test results, lifestyle,
As necessary, clinical data such as information relating to the progress of symptoms, treatment progress, prescription of drugs, specimen management data such as identification information of specimens used for testing, storage conditions, and data related to informed consent of a patient are stored as necessary. . Further, the diagnosis support system of the present invention has a disease-related gene searching means, and the disease-related gene searching means searches for a relationship between the microarray light-emitting pattern and the disease by referring to the microarray light-emitting pattern database and the gene knowledge database. And updates the contents of the disease relevance database based on the search results. The operations of the disease inference means and the disease-related gene searching means are controlled by the control means according to the input from the input / output means.

[0006] As described above, by clarifying the relationship between the microarray emission pattern and the disease, it is possible to accurately estimate a disease that a patient is currently suffering or a disease that is likely to be suffered in the future by using a microarray image. It is possible to provide information that supports the diagnosis by the computer.

[0007]

FIG. 1 is a diagram showing a configuration example of a diagnosis support system according to the present invention. The diagnosis support system of the present invention includes a microarray light emission pattern database 11 for storing a microarray light emission pattern, a disease association database 12 for storing a relationship between the microarray light emission pattern and a disease, and a gene knowledge database for storing knowledge about genes and diseases. A disease inference means for inferring, from the microarray light emission pattern of the patient, a disease that the patient is currently suffering or a disease that is likely to be affected in the future by referring to the microarray light emission pattern database 11 and the disease association database 12; Disease-related gene searching means 15 for searching for a relationship between a microarray light-emitting pattern and a disease by referring to the microarray light-emitting pattern database 11 and the gene knowledge database 13, and inputting and inferring a microarray image And output means 16 for outputting a read microarray image input, the image analysis unit 17 for performing image analysis,
Control means 18 for controlling the operations of the disease inference means 14 and the disease-related gene searching means 15;

The operation of each part of the configuration example of FIG. 1 will be described below in order. The microarray light-emitting pattern database 11 may store a microarray image input from the input / output unit 16 as image data, or perform image analysis by the image analysis unit 17 to calculate a feature amount of the microarray image. The converted data may be stored. Examples of the feature amount include an average value, a maximum value, a minimum value, and a variance of the light emission intensity for each spot of the microarray. When the microarray is composed of N spots, the spots are numbered from 1 to N in advance, and the emission intensity of each spot is stored in the microarray emission pattern database 11 with the emission intensity of the Nth spot as IN. At this time, information such as the type of the microarray used for the test and the name of the gene corresponding to each spot is also stored. The microarray light emission pattern database 11 includes, in addition to information such as the microarray light emission pattern and the type of the microarray, as additional information, basic patient data such as the patient's name, address, date of birth, and family structure, as well as the patient's medical history and family history. Clinical data such as history, chief complaint, findings, test results, lifestyle habits, symptom progression, treatment progress, drug prescription information, sample management data such as sample identification information used for testing, storage status, etc. Stores data related to informed consent as necessary.

Next, an example of a procedure in which the disease-related gene searching means 15 searches for a relationship between a microarray light-emitting pattern and a disease with reference to the microarray light-emitting pattern database 11 and the gene knowledge database 13 will be described. First, the gene knowledge database 13 will be described. The gene knowledge database 13 stores knowledge about genes and diseases. Specifically, information on the sequences and positions (for example, loci) of genes and SNPs, information on the mechanism of the disease, and information related to the disease It stores information on genes and SNPs present, information on genes and SNPs relating to the production of disease-related proteins, etc., and information on know-how in sample storage and analysis. Referring to the microarray light emission pattern database 11 and the gene knowledge database 13, an example of a procedure for searching how the feature amount of each spot of the microarray is involved in the presence or absence of a certain disease is shown in FIG. This will be described with reference to FIG. First, with reference to the microarray luminescence pattern database 11, microarray luminescence patterns of healthy subjects and diseased patients are obtained (procedure 2).
1). If the microarray light emission pattern required for the search is insufficient, it may be newly acquired and stored in the microarray light emission pattern database 11. Next, the microarray light emission patterns of healthy subjects and disease-affected patients were compared to search for characteristic light emission patterns for each (step 2).
2) Identify spots and features effective for classifying healthy subjects and disease-affected patients (step 23). As an example, an example of searching for a relationship between a gene and a disease using cluster analysis will be described. Using the microarray luminescence patterns of healthy subjects and disease-affected patients, the feature amounts of two spots of the microarray (for example, luminescence intensity I1 and I2 at spot 1 and spot 2) are plotted, and clusters are formed using Euclidean distance, Mahalanobis distance, and the like. As a result of the analysis, cluster 1 (31), as shown in FIG.
Three clusters, cluster 2 (32) and cluster 3 (33), are generated, and cluster 2 (32) and cluster 3 (33) are combined to form cluster 4 (3).
4) is generated. At this time, for example, if it is assumed that the cluster 1 (31) matches the healthy group and the cluster 4 (34) matches the diseased group, the result is the emission intensity I
This indicates that the presence or absence of the onset of the disease can be determined based on the magnitude of 1, indicating that the gene corresponding to spot 1 is significantly involved in the onset of the disease. Further, assuming that, of the two clusters generated by dividing the cluster 4 (34) that is the disease affected group, the cluster 2 (32) matches the severe group and the cluster 3 (33) matches the mild group, This result indicates that the severity of the disease can be determined based on the magnitude of the luminescence intensity I2, indicating that the gene corresponding to spot 2 is significantly involved in the severity of the disease. In this manner, one or more spots, ie, genes and their characteristic amounts effective for discriminating between a healthy group and a diseased group, a severe group and a mild group, etc., are identified, and the results are stored in the disease-related database 12. It is stored (procedure 24). When the search results are stored in the disease association database, only medically meaningful results are selected, and confusion does not occur due to erroneous inferences based on medically meaningless knowledge. When searching for an association between SNP and disease, each SN
By associating P with each spot on the microarray, it is possible to perform the same processing as when searching for a relationship between a gene and a disease. FIG. 3 shows an example of a two-dimensional cluster analysis using the feature amounts of two spots for simplicity. However, if the number of spots on the microarray used for search is N, it is expanded to N dimensions. An analysis may be performed. When performing a search, all feature amounts of N spots may be used, or the number of spots N
When is large, the search may be performed after referring to the gene knowledge database 13 to narrow down the candidates of the disease-related gene to some extent in advance. Also, in the example of FIG. 3, cluster analysis is used as the disease-related gene searching means 15, but a statistical method such as multivariate analysis or correlation analysis, a neural network, or the like may be used instead. FIG. 4 shows an example of description contents of the disease association database 12.
The example of FIG. 4 shows a relational database between the disease and the gene corresponding to each spot of the microarray, by quantifying the degree of the emission intensity of each spot of the microarray, that is, the expression level of the gene, for each disease. It is described in the form of In FIG. 4, M types of diseases
(Disease 1, Disease 2, ..., Disease M)
Each disease and N genes (Gene1, Gene2, ...,
GeneN). For example, the numerical value aMN of the cell corresponding to the GeneN column in the row of the disease M indicates the contribution of GeneN to the onset of the disease M. The value of the contribution aMN is represented by a positive value when there is a positive correlation between the expression level of a certain gene and the onset of a disease, and a negative value when there is a negative correlation. In the example of FIG.
When each spot on the microarray corresponds to one gene, the relationship between the gene and the disease is described.
When the relationship between P and the disease is stored in the disease association database 12, it may be described in a similar format by associating each SNP with each spot on the microarray. The disease-related database 12 may store the relationship between the microarray luminescence pattern obtained by the disease-related gene searching means 15 and the disease, or may refer to the gene knowledge database 13 to obtain a known knowledge of genetic medicine. And the relationship between the microarray light emission pattern and the disease may be described and stored.

Next, referring to the microarray light emission pattern database 11 and the disease relation database 12, the disease inferring means 14 is used to determine a disease which the patient is currently suffering or a disease which is likely to be affected in the future from the microarray light emission pattern of the patient. An example of a procedure inferred by the following will be described. FIG. 5 infers from the patient's microarray emission pattern whether the patient is currently suffering from a disease (here, disease M) or whether the patient is likely to suffer from disease M in the future. It is an example of a procedure. First, microarray emission pattern database 1
The characteristic amount of each spot is read as the microarray light emission pattern of the patient from Step 1 (Step 51). Subsequently, the degree of contribution of each spot of the microarray to the disease M is read from the disease relation database 12 (step 52). Based on the data read from the microarray light emission pattern database 11 and the disease association database 12, a procedure 53 is performed.
The sum of the product of the characteristic amount IN of each of the N spots of the microarray and the contribution aMN of GeneN corresponding to each spot to the onset of the disease M is calculated by the equation shown below, and the morbidity index of the patient to the disease M is calculated. calculate. By setting a threshold value for the morbidity index and classifying the morbidity index into several categories according to the magnitude of the morbidity index, the presence or absence of the patient's disease M and the risk of the patient's future disease M being estimated. (Step 54). The processing from step 52 to step 54 is repeated for each disease stored in the disease association database 12 to extract a disease suspected of affecting the patient and a disease likely to be affected by the patient in the future, It is displayed via the input / output means 16.

Next, the operation of the control means 18 will be described. The control means 18 displays on the input / output means 16 whether the disease inference or the disease-related gene search needs to be performed, and activates the disease inference means 14 when the input / output means 16 requests the disease inference. The disease inference means 14 refers to the microarray light emission pattern database 11 and the disease relation database 12 to determine, from the microarray light emission pattern of the patient, a disease that the patient is currently suffering or a disease that is likely to be affected in the future. 1
4 and displayed on the input / output means 16. Also,
When there is a request for searching for a disease-related gene from the input / output means 16 such as when the content of the disease-related database 12 is insufficient, the disease-related gene searching means 15 is operated,
The disease-related gene search means 15 includes a microarray light emission pattern database 11 and a gene knowledge database 1
3, the relationship between the microarray light emission pattern and the disease is searched, and the contents of the disease association database 12 are updated.

[0012]

The diagnostic support system of the present invention accurately estimates a disease that a patient is currently suffering or a disease likely to be caused in the future by comprehensively judging information obtained from a microarray, and Can provide information that assists in diagnosing a wide variety of complex diseases and diseases in which the condition varies from individual to individual. Further, the diagnosis support system of the present invention enables a diagnosis based on the latest knowledge to be performed by searching for a relationship between information obtained from a microarray and a disease and reflecting the obtained result in the system.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration example of a diagnosis support system.

FIG. 2 is a diagram showing an example of a procedure for searching for a relationship between a microarray light emission pattern and a disease.

FIG. 3 is a diagram showing an example of a search result of a relationship between a gene and a disease using cluster analysis.

FIG. 4 is a diagram showing an example of description contents of a disease association database.

FIG. 5 is a diagram showing an example of a procedure for inferring from a microarray light emission pattern whether a patient is currently suffering from a certain disease or is likely to suffer from a certain disease in the future.

[Explanation of symbols]

11: Microarray light emission pattern database, 12
... Disease-related database, 13 ... Gene knowledge database, 14 ... Disease inference means, 15 ... Disease-related gene search means, 16 ... Input / output means, 17 ... Image analysis means, 18 ...
Control means, 21: microarray emission pattern acquisition procedure, 22: characteristic emission pattern search procedure, 23: spot identification procedure for discrimination, 24: search result storage procedure, 31: cluster 1, 32 ... cluster 2, 33 ... cluster 3, 34: cluster 4, 51: feature value reading procedure, 52
... disease contribution reading procedure, 53 ... morbidity index calculation procedure, 54
… Disease determination procedure.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification code FI Theme coat ゛ (Reference) G01N 21/78 G01N 33/53 M 33/53 37/00 102 37/00 102 C12N 15/00 F (72 Inventor Hitoshi Matsuo 1-280 Higashi Koigakubo, Kokubunji-shi, Tokyo Inside the Central Research Laboratory, Hitachi, Ltd. 2G054 AB05 CA22 JA02 JA05 4B024 AA11 HA11 4B029 AA07 AA27 FA15 4B063 QA19 QQ42 QQ52 QR84 QS39 QX02 QX10

Claims (4)

[Claims] 1. A microarray luminescence pattern database for storing a luminescence pattern of a DNA microarray, a disease relation database for storing a relationship between the microarray luminescence pattern and a disease, a gene knowledge database for storing knowledge about genes and diseases, and a microarray. An image analyzing means for analyzing an image, and a disease inference means, wherein the disease inference means refers to the microarray light emission pattern database and the disease relation database, and the patient is currently affected by the microarray light emission pattern. A diagnosis support system characterized by inferring a disease or a disease likely to be affected in the future. 2. The diagnostic support system according to claim 1, wherein the microarray light emission pattern database includes basic patient data such as a patient's name and family structure, a patient's medical history, family history, test results, lifestyle, and medicine. A diagnostic support system for storing medical data such as information on prescriptions. 3. The diagnosis support system according to claim 1, further comprising a disease-related gene searching means, wherein the disease-related gene searching means refers to the microarray luminescence pattern database and the gene knowledge database. And searching for a relationship between the microarray light emission pattern and the disease, and updating the data on the relationship between the microarray light emission pattern and the disease in the disease relation database based on the search result. 4. The diagnosis support system according to claim 1, further comprising an input / output means and a control means, wherein said control means is adapted to receive a disease inference request from said input / output means. A diagnosis support system for activating said disease inference means, and activating said disease-related gene search means when a request for a disease-related gene search is input from said input / output means.