JP2016045801A - Information processor and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To specify a group to which an antibacterial agent belongs in a short period of time.SOLUTION: Generation means 101 generates a determination tree using reaction information showing a reaction when a reagent is given to detected bacteria determined to be drug resistant bacteria among the detected bacteria selected as a sample in a determination test, and classification information showing a result in which the detected bacteria determined to be the drug resistance bacteria is classified into a plurality of groups on the basis of a gene. Extraction means 103 extracts an if-then rule from the determination tree generated by the generation means 101 so as to be stored in a storage part. Acquisition means 104 acquires the reaction information of the detected bacteria different from the sample. Specifying means 105 specifies the classification information corresponding to the if-then rule as the classification information of the detected bacteria from which the reaction information is acquired when the acquired reaction information satisfies the if-then rule from which the acquired reaction information is extracted.SELECTED DRAWING: Figure 3

Description

  The present invention relates to classification of drug-resistant bacteria.

  There are technologies related to classification of drug-resistant bacteria. For example, Patent Document 1 describes a technique for classifying a resistance-resistant bacterium to a group in which it is not possible to deny that the resistance pattern is the same, using the result of determining the sensitivity of a plurality of antibacterial agents by so-called SIR determination. .

JP2013-101675A

  Even drug-resistant bacteria of the same type can be classified into finer groups based on, for example, differences in gene sequences. For example, the drug resistant bacteria classified into a certain group and the drug resistant bacteria classified into another group may have different propagation paths in the medical facility. Therefore, if the group to which the detected drug-resistant bacteria belong can be specified, the propagation path for each group in the medical facility can be specified, which can contribute to the infection control of the drug-resistant bacteria. Therefore, it is important to identify the group to which drug resistant bacteria belong.

In addition to the technique of Patent Document 1, the group to which a drug-resistant bacterium belongs can be determined by analyzing a gene of a drug-resistant bacterium, but the analysis of the gene often takes time. Therefore, it is possible that drug-resistant bacteria are prevalent when analysis results are available.
Therefore, an object of the present invention is to specify a group to which drug-resistant bacteria belong in a short period of time.

  In the determination test for determining whether or not the detected bacterium is a drug resistant bacterium based on a reaction when a reagent is administered to the detected bacterium detected from the specimen, the detected bacterium is determined as the drug resistant bacterium. Means for acquiring reaction information representing the reaction at the time, and the reaction information in the detected bacteria determined as the drug-resistant bacteria in the determination test among the detected bacteria of the sample detected separately from the detected bacteria The classification information corresponding to the reaction information acquired by the acquisition means, using a statistical method using the classification information representing the result of classifying the detected bacteria into a plurality of groups based on genes as objective variables There is provided an information processing apparatus including a specifying unit that specifies

Further, the statistical method is a method using a decision tree generated based on the reaction information and the classification information of the detected bacteria of the sample and a rule extracted from the decision tree,
The specifying means specifies the classification information corresponding to the rule as the classification information corresponding to the acquired reaction information when the acquired reaction information is applied to the extracted rule. Also good.

In addition, the drug-resistant bacterium is MRSA (Methicillin-Resistant Staphylococcus aureus). In the determination test, a drug sensitivity test is performed, and the reaction information in the test may be represented by a MIC (Minimum Inhibitory Concentration) value. .
Furthermore, the analysis of the gene sequence of the detected bacterium is performed using a PFGE (Pulsed-Field Gel Electrophoresis) method, based on the homology probability that the gene sequences of the detected bacterium obtained by the method are homologous. The detected bacteria may be classified into a plurality of groups.

  Further, the creation means for creating the classification information from the analysis result of the gene sequence of the detected bacteria in the sample, the first symbol representing the first group classified by the first homology probability, A second symbol representing a second group classified by a second homology probability higher than a first homology probability, and a third symbol classified by a third homology probability higher than the second homology probability. The code | cord | chord which connected the 3rd symbol showing 3 groups may be provided as a said classification | category information.

  Further, the creating means may determine the fourth homology probability for the classification information representing the group classified by the fourth homology probability that is larger than the second homology probability and smaller than the third homology probability. Thus, a code in which all the third symbols representing the third group to be integrated are connected to the first and second symbols may be created as the classification information.

  In the determination test for determining whether or not the detected bacterium is a drug resistant bacterium based on a reaction when a reagent is administered to the detected bacterium detected from the specimen, the detected bacterium is the drug resistant bacterium. An acquisition means for acquiring reaction information representing the reaction at the time of determination, and the reaction in the detected bacteria determined as the drug-resistant bacteria in the determination test among the detected bacteria of the sample detected separately from the detected bacteria Corresponding to the reaction information acquired by the acquisition means using a statistical method using information as explanatory variables and classification information representing the result of classifying the detected bacteria into a plurality of groups based on genes as objective variables Provided is a program for functioning as a specifying means for specifying the classification information.

  According to the present invention, it is possible to identify a group to which drug-resistant bacteria belong in a short period of time.

The figure showing the whole structure of the information provision system 1 which concerns on embodiment. The figure showing the hardware constitutions of information processor The figure showing the functional composition of an information processor The figure showing an example of the reaction information acquired about a certain detection microbe A figure showing an example of the created classification information Diagram showing an example of the generated decision tree The figure showing an example of the extracted if-then rule The figure showing an example of the sequence of operation | movement of information processing apparatus and a user apparatus

[1] Embodiment [1-1] Configuration FIG. 1 shows an overall configuration of an information providing system 1 according to an embodiment. The information providing system 1 is a system for providing medical information to medical personnel such as doctors and nurses. The information providing system 1 includes a network 2, a user device 3, and an information processing device 10.

  The network 2 is a system that mediates exchange of data between devices, and is, for example, a LAN (Local Area Network) provided in a medical facility. A user device 3 and an information processing device 10 are connected to the network 2. The user device 3 is a device used by a user (mainly a medical worker) of the information providing system 1 and is a computer such as a personal computer, a tablet terminal, or a smartphone. The user device 3 includes display means for displaying an image, and displays information provided in the information providing system 1.

  The information processing apparatus 10 is an apparatus that processes information related to the inspection of microorganisms. Microorganism testing is performed by detecting microorganisms present in specimens such as blood, pus, urine, sputum and pharynx, nasal cavity, tracheal tube, and catheter urine. In particular, the information processing apparatus 10 processes some information (details will be described later) regarding examination of drug-resistant bacteria such as MRSA (Methicillin-Resistant Staphylococcus aureus) and MDRP (Multi-Drug Resistant Pseudomonas). The information processing apparatus 10 processes the information to display information indicating the result of classifying the bacteria detected from the specimen and determined as drug-resistant bacteria into a plurality of finer groups (hereinafter referred to as “classification information”). Is output. By displaying the output classification information on the user device 3, the classification information is provided to the user.

  FIG. 2 shows a hardware configuration of the information processing apparatus 10. The information processing apparatus 10 is a computer that includes a control unit 11, a storage unit 12, and a communication unit 13. The control unit 11 includes a central processing unit (CPU), a read only memory (ROM), a random access memory (RAM), and a real time clock. The CPU controls the operations of the storage unit 12 and the communication unit 13 by executing a program stored in the ROM or the storage unit 12 using the RAM as a work area. The real time clock has a classification function for calculating the current date and time. The storage unit 12 includes a hard disk or the like, and is a storage unit that stores data, programs, and the like used for control by the control unit 11. The communication unit 13 includes a communication circuit for performing communication via the network 2, and transmits and receives data via the network 2.

Functions described below are realized by the control unit 11 of the information processing apparatus 10 executing a program to control each unit.
FIG. 3 shows a functional configuration of the information processing apparatus 10. The information processing apparatus 10 includes a generation unit 101 having a creation unit 102, an extraction unit 103, an acquisition unit 104, an identification unit 105, and an output unit 106.

  The generation unit 101 is a unit that generates a decision tree. The decision tree is also called a decision tree, and is used for determining (specifying in the present embodiment) a second item called an objective variable based on a first item called an explanatory variable. The decision tree is represented by a tree structure that branches for each condition that the first item satisfies.

  In a medical field, a test (hereinafter referred to as “determination”) that determines whether a detected bacterium is a drug-resistant bacterium based on a reaction when a reagent is administered to a bacterium detected from a specimen (hereinafter referred to as “detected bacterium”). Test)). A reagent here is a chemical | medical agent used in order to raise | generate a chemical reaction, and what was prepared for tests, such as an identification test mentioned later and a chemical | medical agent sensitivity test, is used. In the present invention, information (hereinafter referred to as “reaction information”) indicating a reaction when a reagent is administered to a detection bacterium determined as a drug-resistant bacterium in a determination test among detection bacteria selected as a sample is a first matter. That is, it is used as an explanatory variable.

  For example, when MRSA is targeted as a drug-resistant bacterium, the determination test includes a test for identifying Staphylococcus aureus, and a reagent for identifying the drug susceptibility of the detected bacterium. Whether or not the detection bacterium is MRSA is determined based on the reaction when administered and the reaction when a reagent for drug sensitivity is separately administered. Such an MRSA determination test is performed, for example, on a detected bacterium that has been determined to be MRSA based on the shape confirmed with a microscope and the result of Gram staining.

  The drug susceptibility test is performed on the detection bacteria identified as Staphylococcus aureus in the identification test, but may be performed on all detection bacteria that are considered to be MRSA before the result of the identification test is obtained. In the present embodiment, the drug sensitivity test is performed by a method (for example, a micro liquid dilution method) in which the result of the test is represented by a MIC (Minimum Inhibitory Concentration) value. In this test, the MIC value is measured based on the reaction when a reagent for drug sensitivity test is given to the strain. That is, the MIC value is an example of the reaction information described above. Any of the above-described identification test and drug sensitivity test may be performed by a known method.

  Reaction information obtained as a result of the above-described identification test and drug sensitivity test is converted into data, for example, and transmitted to the information processing apparatus 10 via the network 2 shown in FIG. The generation unit 101 acquires the received reaction information and stores it in the storage unit 12. When the reaction information is provided as a printed matter such as paper, the generation unit 101 acquires data obtained by reading the reaction information using an OCR (Optical Character Recognition) technique.

  FIG. 4 shows an example of reaction information acquired for a certain detected bacterium. In FIG. 4A, the reaction information of “A01”, “A02”, “A03”,..., “A40”, which are reagents for the identification test, is “0”, “1”, “0”,. , "1" is shown respectively. “0” represents that there was no reagent reaction, and “1” represents that there was a reagent reaction. In FIG. 4B, reaction information of “Gentamicin”, “Arbekacin”, “Teicoplanin”,..., “Minocycline” as drug sensitivity test reagents is “> 500”, “> 8”, “<= 0”. .5 ”,...,“> 8 ”, respectively.

  Next, in the present invention, information representing the result of classifying the detected bacteria determined to be drug-resistant bacteria into a plurality of groups based on their genes (the above-mentioned “classification information”) is the second matter, Used as an objective variable. In the present embodiment, the sequence of the gene of the detected bacterium is analyzed using a technique called PFGE (Pulsed-Field Gel Electrophoresis). More specifically, from the gene sequence (especially the base pair sequence having a large molecular weight) represented by electrophoresis of a plurality of detection bacteria by this technique, the probability that the genes of the detection bacteria are homologous (hereinafter referred to as “homology probability”). ”), The detected bacteria are classified into a plurality of finer groups. Homology here means that the genes are derived from a common ancestor. For example, when the homology probability is 80%, bacteria having a homology probability of 80% or more are classified as the same group of bacteria. Bacteria classified in the same group in this way have a probability of 80% or more of genes originating from a common ancestor.

  PFGE makes constant the conditions that affect the results of electrophoresis, such as electric field strength, buffer concentration, electroosmosis, and temperature, for a plurality of detection bacteria that are determined to be MRSA by the above determination test. Done. PFGE data indicating the result of the PFGE performed in this way is transmitted to the information processing apparatus 10 via the network 2 shown in FIG. The creation means 102 included in the creation means 101 is means for creating the classification information from the analysis result of the gene sequence of the detected bacteria (in this embodiment, the result of PFGE). The creation unit 102 creates classification information from the PFGE result (that is, the analysis result of the sequence of the gene of the detected bacterium) indicated by the transmitted PFGE data. First, the classification information created by the creation unit 102 will be described with reference to FIG.

FIG. 5 shows an example of the created classification information. In this example, 50 detected bacteria are classified into a plurality of groups. In FIG. 5, for convenience of explanation, numbers from “1001” to “1050” are assigned to these detected bacteria. These are detected bacteria that are detected from specimens collected at either of the hospitals “A” and “O” and determined as MRSA. In FIG. 5, classification information when the homology probabilities of genes are classified as “80%”, “85%”, and “90%” is shown.

  For example, when the homology probability is classified as 80%, “AA1”, “AA2”, “AA3”, “AB12”, “AB3”, “AB4”, “AC1”, “AD1”, “AE1”, “BA123” , “BA4”, “BB12”, and “BB3” are classified into 13 groups. When the homology probability is classified as 85%, “BA123” is divided into “BA1”, “BA2”, and “BA3”, and the detected bacteria are classified into 15 groups. When the homology probability is classified as 90%, “AB12” is divided into “AB1” and “AB2”, “BB12” is divided into “BB1” and “BB2”, and the detected bacteria are classified into 17 groups. .

  In the present embodiment, a plurality of groups into which detected bacteria are classified are respectively represented by combining character strings and numbers. These character strings and numbers are determined based on the following rules. The detection bacteria shown in FIG. 5 are classified into two groups “A” and “B” when the homology probability is 60%. If the homology probability is 70%, the group A is further classified into five groups “A”, “B”, “C”, “D”, and “E”, and the group B is “A”, “B” Are further classified into two groups. Thus, a character string (for example, “AA” or “BA” in which a character string assigned to a group classified as a homology probability of 70% is connected to a character string assigned to a group classified as a homology probability of 60%. Is used as a character string portion of the classification information.

  Subsequently, when the detected bacteria of the group classified as described above are further classified with a homology probability of 90%, “AA” is further classified into three groups “1”, “2”, and “3”. “AB” is further classified into two groups, “1” and “2”. “BA” is further classified into four groups “1”, “2”, “3”, “4”, and “BB” is classified into three groups “1”, “2”, “3”. Further classified. In this way, a code in which a number assigned to a group classified as having a homology probability of 90% is connected after the character string is used as classification information. In addition, a code connected with the number “1” is used as classification information for items such as “AC”, “AD”, and “AE” where the classification does not change even when the homology probability is 90% and 70%. .

  Further, when the homology probability is classified as 85%, if there is a group to be integrated from the case of 90%, a code connecting numbers assigned to the integrated group is used as the classification information. For example, when the homology probability is 90%, the detected bacteria from “1004” to “1029” are classified into “AB1” and “AB2”, but when the homology probability becomes 85%, they are integrated into one group, so “AB12” Is used as classification information. The same applies to the case of classifying the homology probability as 80%. For example, when the homology probability is 90%, the detected bacteria from “1038” to “1044” are classified into “BA1”, “BA2”, and “BA3”, but when the homology probability is 80%, they are integrated into one group. , “BA123” is used as the classification information.

  As described above, in order to create the classification information, the creation unit 102 first creates a first symbol representing the first group classified by the first homology probability (60% in the above example) (the above example). Then, “A” and “B”) are designated. Next, the creating means 102 uses a second symbol (“A” in the above example) representing a second group classified with a second homology probability (70% in the above example) higher than the first homology probability. ”,“ B ”,“ C ”,“ D ”,“ E ”). Then, the creating means 102 uses a third symbol (“1” in the above example) representing a third group classified by a third homology probability (90% in the above example) that is higher than the second homology probability. , “2”, “3”, “4”).

  The creation unit 102 creates a code (for example, “AB1”) connecting the first, second, and third symbols thus designated as classification information. In addition, the creation unit 102 can obtain classification information representing a group classified with a fourth homology probability (80% and 85% in the above example) that is larger than the second homology probability and smaller than the third homology probability. If there is a third group (for example, “AB1” and “AB2”) that is integrated by setting the fourth homology probability, all the third symbols representing the third group are the first and second symbols. A code (for example, “AB12”) connected to is created as classification information.

The generation unit 101 uses the above-described reaction information as the first item (explanatory variable) and the classification information generated by the generation unit 102 as the second item (object variable), for example, a well-known ID3 (Iterative Dichotomiser 3) algorithm or the like. A decision tree is generated using the algorithm.
FIG. 6 shows an example of the generated decision tree. In this example, four branches branch from the root of the decision tree according to the conditions satisfied by the MIC value of “Minocycline”, which is a reagent for the drug sensitivity test. These conditions represent conditions for branching, and are hereinafter referred to as “branch conditions”. The branch “Minocycline = <= 1” is further branched into four branches according to the branching condition satisfied by the MIC value of “Arbekacin”.

  The first branch indicates that the classification information “BB2” is specified when “Arbekacin = 4”. The numerical value “4-> 2” shown after the classification information indicates that there are four detected bacteria satisfying the branching conditions so far, and two of them are classified as “BB2”. In addition, when “Arbekacin = 2”, it is further branched into two branches, and when the branch condition “Teicoplanin = <= 0.5” is satisfied, the classification information “BA4” is specified. It is shown. When the generation unit 101 generates a decision tree, the generation unit 101 supplies the generated determination tree to the extraction unit 103.

  The extraction unit 103 is a unit that extracts a rule for identifying classification information from the decision tree generated by the generation unit 101. The rule extracted from the decision tree is a rule expressed in an if-then format in the present embodiment, and is hereinafter referred to as an if-then rule. When the first item is applied to the if-then rule, the second item corresponding to the if-then rule is specified. Note that the second item specified in this way is not necessarily 100% accurate, but the extraction unit 103 extracts a rule that specifies the second item as accurately as possible as an if-then rule. When the decision tree shown in FIG. 6 is supplied from the generation unit 101, the extraction unit 103 extracts the following if-then rule.

  FIG. 7 shows an example of the extracted if-then rule. In this example, the second item (classification information such as “BB1” and “AE1”) represented in the decision tree shown in FIG. 6 is defined as the “then” part, and the branch connecting the root of the decision tree and the “if” part is defined as the “if” part. An if-then rule is represented. Specifically, when the reaction information which is the first item is “Minocycline = <= 1”, “Arbekacin = 2” and “Teicoplanin = 1” as shown in the if part, the second information The if-then rule that the classification information as a matter is specified as “BA1” as shown in the “then” part is shown.

  “Probability: 100.0%” shown in FIG. 7 means that when the if-then rule is applied in the decision tree, the classification information is BA1 with a probability of 100.0%. doing. As another example, if the reaction information is “Minocycline => 8”, “Arbekacin = 2” and “Gentamicin => 500” as shown in the if part, the classification information is shown in the then part. As shown in FIG. 7, the if-then rule (probability is 66.6%) is specified as “AB3”.

  When the extracting unit 103 extracts the if-then rule from the decision tree using a known technique, the extracting unit 103 stores the rule data representing the extracted if-then rule in the storage unit 12 of the own device. Specifically, the rule data is data representing the above-described “if” part and “then” part. The generation of the decision tree and the extraction of the if-then rule are performed as a preparation stage before the information providing system 1 provides the classification information to the user.

  For the detection bacteria selected as the above-mentioned sample, in the above preparation stage, in addition to the drug-resistant bacteria determination test, classification information was created based on the gene sequence analysis results. After the rule is stored, the classification information is specified using only the reaction information of the determination test without performing PFGE. This specification is triggered by, for example, request data indicating that the classification information of the detected bacteria is requested from the user apparatus 3 to the information processing apparatus 10 together with the reaction information of the detected bacteria subjected to the determination test. Done.

The acquisition means 104 obtains reaction information indicating a reaction of a reagent when a detected bacterium detected from a sample different from the sample from which the detected bacterium selected as a sample is detected as a drug-resistant bacterium in the determination test. get. For example, the acquisition unit 104 receives the request data described above, and acquires reaction information indicated by the received request data. In addition to this, for example, when the reaction information of the detected bacteria subjected to the determination test is obtained, it may be stored in the storage unit 12 of the information processing apparatus 10. In this case, when the user operates the user device 3 to select the reaction information stored in the information processing device 10 and requests specification of the classification information, the acquisition unit 104 performs classification from the storage unit 12 of the own device. The reaction information requested for information may be acquired from the storage unit 12. The acquisition unit 104 supplies the acquired reaction information to the specifying unit 105.

  The identifying unit 105 identifies the classification information corresponding to the reaction information acquired by the acquiring unit 104 using a statistical method using the reaction information as an explanatory variable and the classification information as an objective variable. In this embodiment, the identification unit 105 uses a method using a decision tree generated based on the reaction information and classification information of the detected bacteria in the sample and a rule extracted from the decision tree as the statistical method. . When the reaction information is supplied from the acquisition unit 104, the specifying unit 105 reads out the if-then rule (the if-then rule extracted by the extraction unit 103) by reading out the rule data stored in the storage unit 12. .

  The specifying unit 105 determines whether or not the reaction information supplied from the acquiring unit 104 is applicable to the read if-then rule. When determining that it is true, the specifying unit 105 specifies the classification information represented by the “then” part in the “if-then” rule as the classification information corresponding to this “if-then” rule. As described above, when the reaction information acquired by the acquiring unit 104 is applied to the if-then rule extracted from the decision tree, the specifying unit 105 acquires the classification information corresponding to the if-then rule. Specified as classification information corresponding to the reaction information. The specifying unit 105 supplies the specified classification information to the output unit 106.

  The output unit 106 is a unit that outputs the classification information specified by the specifying unit 105 to a predetermined destination. When the classification information requested from the user device 3 is specified as described above, the output unit 106 outputs the classification information with the user device 3 as a destination. In addition, the output unit 106 may output the classification information with the display unit as a destination when the own apparatus includes the display unit, or the user's e-mail address or SNS (Social Networking Service) account. Classification information may be output as a destination. As described above, the output unit 106 outputs the classification information to a destination determined to transmit the identified classification information to the user.

[1-2] Operation The information providing system 1 performs a classification information providing process for providing classification information to a user based on the above configuration. Below, the operation | movement which each apparatus with which the information provision system 1 is provided in a classification information provision process is demonstrated with reference to FIG.
FIG. 8 illustrates an exemplary sequence of operations of the information processing apparatus 10 and the user apparatus 3. First, the information processing apparatus 10 acquires reaction information obtained for detected bacteria (both MRSA) (step S11). Next, the information processing apparatus 10 acquires the analysis result of the genes of those detected bacteria (analysis result by PFGE in this embodiment) (step S12). The reaction information in step S11 is acquired about two days after the detected bacteria are detected, but it takes about one week until the analysis result in step S12 is acquired.

  Subsequently, the information processing apparatus 10 creates classification information using the acquired analysis result (step S13). The creation unit 102 performs the operation in step S13. And the information processing apparatus 10 produces | generates the decision tree using the acquired reaction information and the produced classification information (step S14). The generation unit 101 performs the operation in step S14. Next, the information processing apparatus 10 extracts the if-then rule from the generated decision tree (step S15) and stores it in the storage means (the storage unit 12 of the own apparatus in the present embodiment) (step S16). The extraction means 103 performs the operations of steps S15 and S16.

  Thereafter, for example, the user device 3 acquires reaction information about the detected bacteria detected from the new specimen (detected bacteria determined as MRSA in the determination test) (step S21). When the user performs an operation for requesting the classification information of the detected bacteria and receives the operation (step S22), the user device 3 transmits request data indicating the request to the information processing device 10 (step S23). When receiving the request data, the information processing apparatus 10 acquires, for example, reaction information indicated by the request data (step S24). The acquisition unit 104 performs the operation in step S24.

  Subsequently, the information processing apparatus 10 identifies classification information based on the acquired reaction information and the if-then rule stored in step S16 (step S25), and outputs the identified classification information to the user apparatus 3. (Step S26). The specifying unit 105 performs the operation in step S25, and the output unit 106 performs the operation in step S26. Upon receiving the output classification information, the user device 3 displays it (step S27). In this way, the user can confirm the requested classification information.

[1-3] Effect In this embodiment, the if-then rule extracted from the decision tree using the reaction information of the detected bacteria as explanatory variables and the classification information based on the genes of the detected bacteria as the objective variable is displayed in the information processing apparatus 10. Once stored, if there is reaction information of the detected bacteria, then the classification information of the detected bacteria is specified. As described above, the period required for obtaining the reaction information is about two days, but the period required for analyzing the gene sequence by PFGE is about one week. That is, according to the present embodiment, the classification information is specified only from the reaction information, so that the group to which the drug-resistant bacteria belong is compared with the case where the classification information is created based on the analysis result of the gene sequence. The identification can be performed in a short period of time, and as a result, the classification information representing the identified group can be provided to the user in a short period of time.

  In this embodiment, MRSA is targeted as a drug resistant bacterium, and MIC value is used as reaction information. As a result, for example, in the decision tree shown in FIG. 6, the same classification information as the classification information representing the group classified based on the analysis result of the gene sequence was identified for 39 detected bacteria out of 50 detected bacteria. . That is, accurate classification information is specified with a probability of 39 ÷ 50 = 0.78, that is, 78%. This probability was about 60% when a decision tree generated using only the reaction information in the identification test was used. Thus, according to the present embodiment, by using both the reaction information in the identification test and the reaction information in the drug sensitivity test, accurate classification information can be obtained with higher probability than when only the reaction information in the identification test is used. Can be identified.

  In this embodiment, classification information, that is, information indicating the result of classifying the detected bacteria determined to be drug-resistant bacteria based on the gene is used as the objective variable. For example, from the PFGE result, the genes of the detected bacteria It is impossible to use it as an objective variable just by knowing the homology probability of. In the present embodiment, classification information can be used as an objective variable by using, as classification information, a group that is classified based on a homology probability as described above and represented by a code.

  In particular, in this embodiment, when the homology probability is the above-described third homology probability (90% in the example of FIG. 5), the classification information is a code connecting the first, second, and third symbols. Classification information was presented. Thereby, when there are two drug-resistant bacteria, it is possible to know whether or not the bacteria are classified into the same group when the homology probability is lowered only by looking at their classification information. For example, if “AB1” and “AC1” are the same, the first symbol is the same and the second symbol is different. Therefore, if it is reduced to the first homology probability (60% in the example of FIG. 5), it is classified into the same group. I understand that In addition, if “AB1” and “AB2”, it is the same up to the second symbol, so it can be seen that they are classified into the same group simply by lowering to the second homology probability (70% in the example of FIG. 5). .

  Further, in the present embodiment, the classification information created with the fourth homology probability (80% and 85% in the example of FIG. 5) includes the third group that is classified separately if it is the third homology probability. All the third symbols to represent are included. For example, in the example of FIG. 5, the classification information “BA123” includes three third symbols “1”, “2”, and “3” included in “BA1”, “BA2”, and “BA3”, respectively. It is included. From this relationship, just by looking at the classification information “BA123” created with the fourth homology probability, the drug-resistant bacteria classified into the group represented by the classification information are classified into three separate groups with the third homology probability. You can know that it is classified.

[2] Modification The above-described embodiment is merely an example of the present invention and may be modified as follows. Moreover, you may implement combining embodiment mentioned above and each modification shown below as needed.

[2-1] Update of if-then rule The once-then rule once extracted may be updated. For example, in the embodiment, a decision tree is generated from reaction information and classification information obtained from 50 detected bacteria and if-then rules are extracted, but reaction information and classification information obtained from other detection bacteria are added. The if-then rule is updated by generating the determined tree and extracting the if-then rule. Thus, as the parameters of the first and second items used in the decision tree increase, the possibility of specifying accurate classification information can be increased. However, in that case, it is desirable that the conditions (field strength, buffer solution concentration, electroosmosis, temperature, etc.) for performing PFGE to analyze the gene sequence are the same as much as possible.

[2-2] Classification information The classification information is not limited to that represented in the embodiment. For example, each of the first, second, and third symbols, such as “AA1” and “AB1”, may be a plurality of characters or numbers instead of a single character or number. Moreover, you may represent classification information by connecting 4 or more symbols. In addition to letters and numbers, symbols representing operators such as “+” and “−” and figures such as “Δ” and “□” may be used. More specifically, in the case of MRSA, a name assigned to each group of bacteria such as “USA100” and “USA300” may be used as the classification information.

[2-3] Drug Sensitivity Test In the embodiment, a drug sensitivity test whose result is represented by an MIC value was performed, but the present invention is not limited to this. The drug sensitivity test may be performed by a known method such as a dilution method or a diffusion method as long as the above-described reaction information (information indicating a reaction by a reagent) can be obtained. If a new method is found in the future, the drug sensitivity test may be performed. It may be performed by a method. Note that the degree of accuracy of the classification information specified by the extracted if-then rule differs depending on the test method. Therefore, it is desirable to adopt a method in which classification information is specified more accurately.

[2-4] Classification Based on Genes In the embodiment, the detection bacteria are classified into a plurality of groups by analyzing the sequences of the genes of the detection bacteria using PFGE. However, the present invention is not limited to this. For example, the analysis may be performed by the POT method (Phage Open Reading Frame Typing method), or the analysis may be performed by another method. In the information providing system 1, classification may be performed by any method as long as it is a method that can classify a group of drug-resistant bacteria with higher accuracy than the above-described test for drug-resistant bacteria. The higher the accuracy of classification of drug-resistant bacteria, the more accurate the classification information specified by the extracted if-then rule.

[2-5] Drug-resistant bacteria The embodiment has been described by taking the case where MRSA is classified into a plurality of groups as an example. However, even with MDRP, a drug sensitivity test result can be obtained as a MIC value. Classification can be performed similarly to the embodiment. In addition to these, for example, classification information of drug resistant bacteria such as ESBL (Extended Spectrum Beta-Lactamase) and multidrug resistant acinetobacter may be specified. In this case, if a decision test and gene sequence analysis that can be performed are performed and a decision tree is generated from the reaction information and classification information obtained as a result, the if-then rule is extracted as in the embodiment. Classification information can be specified.

[2-6] Reaction Information In the embodiment, the reaction information obtained in both the identification test and the drug sensitivity test is used, but one of the reaction information may be used. In that case, compared to the case where both reaction information is used, the probability that accurate classification information is specified may be reduced. However, the classification information of drug-resistant bacteria can be specified in a short period of time. It is not different from the form. In addition, as described above, it is conceivable that the probability is increased by changing the drug sensitivity test method or the drug-resistant bacteria to be classified. Therefore, it is not necessary to limit that the response information in both the identification test and the drug sensitivity test is essential.

[2-7] Homology Probability In the embodiment, 60%, 70%, 90%, 80%, and 85% are used as the first, second, third, and fourth homology probabilities, but the present invention is not limited to these. Instead, a larger value may be used as each homology probability, or a smaller value may be used. In any case, it is sufficient that the first homology probability <the second homology probability <the fourth homology probability <the third homology probability. However, when the first homology probability, the second homology probability, the fourth homology probability, the third homology probability, and the homology probability are changed, the group to be classified is larger than the change does not change much. It is better to change. This is because the hierarchical structure of the group of drug resistant bacteria becomes clearer.

[2-8] Processing in External Device The processing performed by the generation unit 101, the creation unit 102, and the extraction unit 103 may be performed by the external device instead of the information processing device 10. In that case, the external apparatus creates classification information, generates a decision tree, and extracts an if-then rule, and transmits rule data representing the extracted if-then rule to the information processing apparatus 10. The specifying unit 105 of the information processing apparatus 10 specifies the classification information based on the if-then rule indicated by the transmitted rule data.

[2-9] Statistical Method In the embodiment, the identifying unit 105 uses the decision tree and the if-then rule as the statistical method, but the present invention is not limited to this. In addition, the identifying unit 105 may identify classification information from the acquired reaction information after learning using the reaction information and classification information of the detected bacteria of the sample using, for example, a neural network technique. Good. In addition to this, any statistical method may be used as long as it is a method for identifying an objective variable from other explanatory variables based on a result of statistical analysis of given explanatory variables and objective variables. .

[2-10] Category of Invention The present invention can be understood as an information providing system including the information processing apparatus and the user apparatus. Further, the present invention can be understood as an information processing method for realizing processing performed by the information processing apparatus or the user apparatus, or as a program for causing a computer such as the information processing apparatus or the user apparatus to function. . This program may be provided in the form of a recording medium such as an optical disk in which it is stored, or may be provided in the form of being downloaded to a computer via a network such as the Internet, installed and made available for use. May be.

DESCRIPTION OF SYMBOLS 1 ... Information provision system, 2 ... Network, 3 ... User apparatus, 10 ... Information processing apparatus, 11 ... Control part, 12 ... Memory | storage part, 13 ... Communication part, 101 ... Production | generation means, 102 ... Creation means, 103 ... Extraction means , 104 ... acquisition means, 105 ... identification means, 106 ... output means

Claims (7)

The reaction when the detected bacterium is determined to be the drug-resistant bacterium in a determination test for determining whether or not the detected bacterium is a drug-resistant bacterium based on the reaction when the reagent is administered to the detected bacterium detected from the sample. Obtaining means for obtaining reaction information representing
Among the detected bacteria of the sample detected separately from the detected bacteria, the reaction information in the detected bacteria determined as the drug-resistant bacteria in the determination test is used as an explanatory variable, and the detected bacteria are grouped into a plurality of groups based on genes. An information processing apparatus comprising: a specifying unit that specifies the classification information corresponding to the reaction information acquired by the acquisition unit using a statistical method using classification information representing the classification result as an objective variable. The statistical method is a method using a decision tree generated based on the reaction information and classification information of the detected bacteria of the sample and a rule extracted from the decision tree,
The specifying means specifies the classification information corresponding to the rule as the classification information corresponding to the acquired reaction information when the acquired reaction information is applied to the extracted rule. Item 4. The information processing apparatus according to Item 1. The drug resistant bacterium is MRSA (Methicillin-Resistant Staphylococcus aureus),
The information processing apparatus according to claim 1, wherein a drug sensitivity test is performed in the determination test, and the reaction information in the test is represented by a MIC (Minimum Inhibitory Concentration) value. The analysis of the gene sequence of the detected bacteria is performed using a PFGE (Pulsed-Field Gel Electrophoresis) technique, and the detection is performed based on the homology probability that the gene sequences of the detected bacteria obtained by the technique are homologous. The information processing apparatus according to any one of claims 1 to 3, wherein the bacteria are classified into a plurality of groups. A creation means for creating the classification information from the analysis result of the gene sequence of the detected bacteria in the sample, the first symbol representing the first group classified by the first homology probability, and the first A second symbol representing a second group that is classified with a second homology probability that is higher than the second homology probability, and a third symbol that is classified with a third homology probability that is higher than the second homology probability. The information processing apparatus according to claim 4, further comprising a creation unit that creates a code connecting a third symbol representing a group as the classification information. The creation means sets the fourth homology probability for the classification information representing the group classified by the fourth homology probability that is larger than the second homology probability and smaller than the third homology probability. The information processing apparatus according to claim 5, wherein a code in which all of the third symbols representing the third group integrated in step 1 are connected to the first and second symbols is created as the classification information. Computer
The reaction when the detected bacterium is determined to be the drug-resistant bacterium in a determination test for determining whether or not the detected bacterium is a drug-resistant bacterium based on the reaction when the reagent is administered to the detected bacterium detected from the sample. Obtaining means for obtaining reaction information representing
Among the detected bacteria of the sample detected separately from the detected bacteria, the reaction information in the detected bacteria determined as the drug-resistant bacteria in the determination test is used as an explanatory variable, and the detected bacteria are grouped into a plurality of groups based on genes. A program for functioning as a specifying means for specifying the classification information corresponding to the reaction information acquired by the acquisition means, using a statistical method using classification information representing the classification result as an objective variable.

Images