Method and system for interpreting and validating experimental data with automated reasoning

• US 6,813,615 B1
• Filed: 09/06/2000
• Issued: 11/02/2004
• Est. Priority Date: 09/06/2000
• Status: Expired due to Term

• Current Assignee: Cellomics Incorporated (Thermo Fisher Scientific Incorporated)
• Original Assignee: Cellomics Incorporated (Thermo Fisher Scientific Incorporated)
• Inventors: Collins, Mark A. D., Shaw, John R., Colasanti, Ricardo L.
• Primary Examiner(s): Starks, Jr., Wilbert L.
• Assistant Examiner(s): Hirl, Joseph P.
• Application Number:
  US09/656,372
• Time in Patent Office: 1,518 Days
• Field of Search: 706/45, 706/46, 706/47, 704/9
• US Class Current: 706/46
• CPC Class Codes: G06F 2216/03 Data mining
• CPC Class Codes: G06N 5/045 Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
• CPC Class Codes: G16C 20/70 Machine learning, data mining or chemometrics
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• CPC Class Codes: G16H 70/40 relating to drugs, e.g. their side effects or intended usage

Abstract

A method and system for interpreting experimental data with automated reasoning. Domain specific knowledge is acquired from one or more pharmaceutical information sources. A semantic representation of the domain specific knowledge is created meeting a desired set of criteria. Pharmaceutical data from a knowledge database is classified with the semantic representation, allowing construction of a set of reasons for any classified pharmaceutical data. The set of reasons may help interpret the classified pharmaceutical data to remove errors, such as “physical errors” and “biological errors”. Removing such errors helps improve fusion of knowledge from multiple data, information and knowledge sources which incorporates activity and selectivity against a target, desired pharmacokinetic and toxicity properties enabling selection of potential pharmaceutical compounds. The method and system may improve identification, selection, validation and screening of new real or virtual pharmaceutical compounds or may be used to provide bioinformatic techniques for storing and manipulating pharmaceutical knowledge.

Claims

1. A method for interpreting experimental data with automated reasoning, comprising:

acquiring domain specific knowledge from a plurality of pharmaceutical information sources;

creating a semantic representation of the domain specific knowledge that meets a desired set of criteria;

classifying pharmaceutical data from a knowledge database with the semantic representation;

providing a set of reasons for any classified pharmaceutical data, wherein the set of reasons are used to help interpret the classified pharmaceutical data;

creating a further semantic representation of the domain specific knowledge;

classifying pharmaceutical data from the knowledge database with the further semantic representation; and

creating fused knowledge from the classified pharmaceutical data.

2. A computer readable medium having stored therein instructions for causing a processor to execute the method of claim 1.

3. The method of claim 1 wherein the fused knowledge includes knowledge obtained from a plurality of domains from pharmaceutical industries fused into a multi-parameter output in a single parallel pass through the knowledge database.

4. The method of claim 1 wherein the step of creating a semantic representation of the domain specific knowledge that meets a desired set of criteria includes creating a semantic representation of a general screening expert, an instrument expert or an assay expert using a plurality of expert specific rules.

5. The method of claim 1 wherein the step of classifying pharmaceutical data from a knowledge database with semantic representation includes classifying pharmaceutical data based on determined physical errors from a screening process used to collect the pharmaceutical data.

6. The method of claim 5 wherein the physical errors include gel-electrophoresis errors, bio-chip errors, pipettor errors, microplate preparation errors or microplate variance errors.

7. The method of claim 1 wherein the step of classifying pharmaceutical data from a database with semantic representation includes classifying pharmaceutical data based on determined biological errors from an assay used to collect the pharmaceutical data.

8. The method of claim 1 wherein the step of providing a set of reasons for any classified pharmaceutical data includs providing a set of reasons as to why a detected pattern in the classified pharmaceutical data is an error pattern.

10. The method of claim 5 wherein the step of creating a semantic representation of the domain specific knowledge that meets a desired set of criteria includes creating a semantic representation of a general screening expert, an instrument expert or an assay expert using a plurality of expert specific rules.

11. The method of claim 5 wherein the step of classifying pharmaceutical data from a knowledge database with semantic representation includes classifying pharmaceutical data based on determined physical errors from a screening process used to collect the pharmaceutical data.

12. The method of claim 11 wherein the physical errors include gel-electrophoresis errors, bio-chip errors, pipettor errors, microplate preparation errors or microplate variance errors.

13. The method of claim 5 wherein the step of classifying pharmaceutical data from a database with semantic representation includes classifying pharmaceutical data based on determined biological errors from an assay used to collect the pharmaceutical data.

14. The method of claim 5 wherein the step of providing a set of reasons for any classified pharmaceutical data includes providing a set of reasons as to why a detected pattern in the classified pharmaceutical data is an error pattern.

9. A method for interpreting experimental data with automated reasoning, comprising:

acquiring domain specific knowledge from a plurality of pharmaceutical information sources;

creating a semantic representation of the domain specific knowledge that meets a desired set of criteria;

classifying pharmaceutical data from a knowledge database with the semantic representation;

providing a set of reasons for any classified pharmaceutical data, wherein the set of reasons are used to help interpret the classified pharmaceutical data;

determining with the set of reasons whether any classified pharmaceutical data includes data related to physical errors or biological errors, and if so,

marking classified pharmaceutical data related to physical errors or biological errors

as unreliable in the knowledge database, thereby validating any fused knowledge

created from the knowledge database.

15. A method for interpreting experimental data with automated reasoning, comprising:

acquiring domain specific knowledge from a plurality of pharmaceutical information sources;

creating a semantic representation of the domain specific knowledge that meets a desired set of criteria, wherein the semantic representation includes plurality of rules to identify physical errors or biological errors in a plurality of screening processes used to collect pharmaceutical data;

classifying a plurality of errors patterns in pharmaceutical data from a knowledge database with the semantic representation;

providing a set of reasons for any classified pharmaceutical data, wherein the set of reasons are used to annotate error patterns to help interpret physical errors in the classified pharmaceutical data; and

marking the classified pharmaceutical data as unreliable in the knowledge database, thereby validating any fused knowledge created from the knowledge database, wherein the fused knowledge includes knowledge obtained from a plurality of domains from pharmaceutical industries fused into a multi-parameter output in a single parallel pass through the knowledge database.

16. A computer readable medium having stored therein instructions for causing a processor to execute the method of claim 15.

17. The method of claim 15 wherein the physical errors include gel-electrophoresis errors, bio-chip errors, pipetter errors, microplate preparation errors or microplate variance errors and biological errors including assay errors.

18. The method of claim 15 wherein the step of creating a semantic representation of the domain specific knowledge that meets a desired set of criteria includes creating a semantic representation of a general screening expert, an instrument expert or an assay expert using a plurality of expert specific rules.

19. An automated reasoning creation and analysis system, comprising in combination:

an automated reasoning engine for acquiring domain specific knowledge from a plurality of pharmaceutical information sources, creating a semantic representation of the domain specific knowledge that meets a desired set of criteria, classifying pharmaceutical data from a knowledge database with the semantic representation, and providing a set of reasons for any classified pharmaceutical data, wherein the set of reasons are used to help interpret the classified pharmaceutical data, creating a further semantic representation of the domain specific knowledge, classifying pharmaceutical data from the knowledge database with the further semantic representation, determining with the set of reasons whether any classified pharmaceutical data includes data related to physical errors or biological errors, and if so, marking classified pharmaceutical data related to physical errors or biological errors as unreliable in the knowledge database, thereby validating any fused knowledge created from the knowledge database, and wherein the fused knowledge includes knowledge obtained from a plurality of domains from pharmaceutical industries fused into a multi-parameter output in a single parallel pass through the knowledge database;

plurality of domain specific knowledge from a plurality of pharmaceutical information sources; and

a knowledge database for storing raw experimental data and knowledge derived from raw pharmaceutical data.

20. The system of claim 19 wherein creating semantic representation includes creating a semantic representation of a general screening expert, an instrument expert or an assay expert using a plurality of expert specific rules.