JP2004500614A - Receptor selectivity mapping - Google Patents

Abstract

A computer system comprising a first database containing records about compounds and the effects of the compounds on biological systems of humans and animals, and a second database containing records about molecular targets. The system further includes a third database containing records relating to binding, reactivity and other interaction tests between the compounds of the first database and the molecular targets of the second database. The test includes information on the effects of the compound in the interaction between the specific molecular target in the second database and the compound known to interact with the specific molecular target. Means for setting the test threshold and for selecting compounds that meet the interaction test threshold are also included in the system. A user interface is provided that allows a user to browse, operate, or analyze information in the first, second, and third databases.
[Selection diagram] FIG.

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention generally relates to a technique for constructing a multidimensional database by combining chemical informatics and bioinformatics with data on the interaction of chemical molecular targets. In particular, the present invention relates to a database containing compounds, molecular targets, and biological or clinical information, in which patterns or relationships of interactions between the compounds and molecular targets are determined and stored in a database. Comparisons with other information have led to conclusions that will be useful in drug discovery and development and related areas.
[0002]
[Prior art]
The global pharmaceutical industry spends $ 30 billion a year on research and development, of which approximately one-third is the time it takes to select drug candidates for preclinical and clinical development. Spent during the early stages of development. The important clinical steps for drug discovery consist of the following steps. (1) a sequence of DNA containing a segment of the human genome, (2) identification of a gene having a genome associated with a particular disease or biological function, (3) associated with or by the functional gene. Generation of proteins such as receptors or enzymes that are encoded and later become biological or molecular targets for drug discovery, (4) screening from compound libraries for activity against molecular targets, (5) other biology To screen for compounds that are most active against a specific target, assess the selectivity or specificity of the compound for the biological / molecular target of interest, and the potential for unwanted side effects due to activity against other targets (6) To test characteristics such as toxicity, absorption, distribution, metabolism, and excretion (7) The most promising compound is evaluated based on empirical judgment using the above information, and the information is sent to the chemical synthesis group. (8) After retesting the chemical analogs in steps (4), (5) and (6) until the optimal derivative or group of compounds is identified (8) Repeat 7), (9) Use this optimal derivative for further preclinical and clinical trials.
[0003]
In this discovery and development process, compounds that pass through narrow filters are screened for more expensive preclinical and clinical development. Unfortunately, in preclinical and clinical development that follows this screening process, compounds often fail to pass these stages and do not reach commercialization. Due to these failures, the average cost of developing and launching one new drug is estimated to be over $ 300 million. However, if the best drug candidates can be accurately identified early in the discovery and development process, and if the drug clears preclinical and clinical trials, development costs can be reduced by as much as 75%. Can be. Clearly, a major objective in pharmaceutical research and development (R & D) is to improve the predictability of such early drug development tests.
[0004]
Innovations in biotechnology and the development of instruments that can automate many of the experimental processes have created two major trends that have a significant impact on pharmaceutical R & D. First, due to advances in the sequencing of the human genome, the number of molecular targets (such as new receptors and enzymes) that can be used in new drug discovery screening programs continues to increase rapidly. Approximately 400 molecular targets are being investigated for new drug discovery, and the number of potential molecular targets that may be elucidated by the human genome project is estimated to be in the thousands to more than 10,000. Second, new technologies such as automation and combinatorial chemistry have increased the size of compound libraries that can be used in new drug discovery screening programs by approximately 10-fold (more than one million compounds found in many pharmaceutical companies). ) Is expanding. While these two factors offer great promise for new drug discovery, they also create significant potential problems with undesirable cost consequences for new drug development. More targets and compounds lead to the discovery of additional bioactive compounds, resulting in greater difficulties in selecting the best drug candidate to proceed to preclinical trials, and more compounds are needed for preclinical trials and Development costs will increase as going to clinical trials creates more failures at these stages.
[0005]
These factors increasingly require rapid and low cost ("tube" or microplate-based) in vitro assays in the selection, optimization and detection of derived compounds. Such a rapid assay would help identify the most promising of these active compounds before moving on to later expensive drug development stages. These factors further require more effective methods for managing and interpreting vast amounts of data on genes and gene products (molecular targets), chemical structures, and screening results.
[0006]
One application of in vitro assays, which has become increasingly important in pharmaceutical R & D, is "profiling." The patentee of this patent application pioneered the concept of profiling in the late 1980s. Pharmaceutical companies have a vast array of in vitro assays to characterize the pharmacological activity and potential side effects of compounds developed as new drugs. Currently, molecular targets play important roles in a wide range of human diseases, including diseases involving central nervous system disorders, immune diseases, pain and inflammation, infectious diseases, cancer, metabolic or growth factors, cardiovascular function, and endocrine system There are over 200 different assays that are routinely performed, based on the required receptors and enzymes. Drugs account for more than half of the global market function due to their interaction with cell receptors. In addition, the side effects of many drugs have been mitigated by their interaction with receptors and enzymes.
[0007]
In profiling, derived compounds of a pharmaceutical company, which are generally in the preclinical development phase, are tested with receptors and enzyme assay devices. Information on the interaction of this company's compounds with specific receptors, obtained from the profiling process, is important in the optimization and selection of derived compounds and suggests any side effects or potential second effects of the compounds. Become. With this knowledge, the pharmaceutical company can potentially save millions of dollars in the time and cost of preclinical and / or clinical development of the compound.
[0008]
Although profiling services have been available for many years, pharmaceutical companies have typically used data from these tests empirically. Many drugs interact with many receptors or other molecular targets, including highly selective drugs. Therefore, the data generated by profiling should be interpreted by researchers at the pharmaceutical company based on experience and knowledge, taking into account both the data on the chemical structure of the compound and the binding activity of the compound to a particular receptor. It is. Unfortunately, even the most experienced pharmacologists have incomplete knowledge of the interaction of various drugs with a wide range of receptors associated with new drug development.
[0009]
The need for more effective methods to manage, collate, interpret and utilize vast amounts of data on genes and gene products (molecular targets), chemical structures, and screening results has led to bioinformatics and New opportunities have been created in chemical informatics or the management of biological and scientific data. The steps to creating a vast pool of information for new drug discovery consist of the following steps: (1) DNA sequencing (coding of genetic material or gene that serves as a blueprint for cells to produce gene products or proteins), (2) Functional genomics (particularly in response to changes in drug or biological function) , The process of converting a DNA sequence into the relevant gene product or protein via the mRNA product), (3) proteomics (amino acid sequence and / or gene product such as a receptor encoded in the gene or (4) Identification of three-dimensional structure of protein), (4) Pharmacology / toxicology of trace molecules (molecular binding or interaction between gene products such as receptors and small organic compounds that can be drugs), (5) Chemical Structure (for micromolecules, drug analogs).
[0010]
Databases for DNA sequencing (Group 1) have been established and include Genbank, Genome Center and others. Similarly, chemical structure databases (Group 5) are also well known and are provided by vendors such as MDL (Isis) and Oxford Molecular. Proteomics databases (Group 3), such as SWISS-PROT, ProLink and PDB, have also been constructed. These databases contain structural information and can be used to determine patterns in one dimension or in one component of structural or sequence information, so each can be considered as one component. . The databases of Groups 2 and 4 are not well established yet, but will provide valuable additional information in the information pool for drug discovery and development. These latter two types of databases contain data on the interaction between the two structures, such as gene-to-protein (group 2) and protein-to-compound (group 4). And it is two-dimensional. Such a database relationship has an added level of complexity compared to a database consisting of one component.
[0011]
Partial databases or multiple databases for Group 4 protein-to-compound relationships are currently being constructed. For example, the binding profile of a single compound to a wide range of receptor targets provided by the assignee to the client is a partial dataset of a group 4 type database. Similarly, advanced processing such as screening thousands to hundreds of thousands of compounds, such as those contained in a chemical structure database (Group 5), for activity against a particular receptor target (a single point in Group 3) The data generated by the screening project will be part of the Group 4 database. While such partial group 4 databases would be useful in new drug discovery and development, they have two major drawbacks. First, they are specific components of two components, such as the binding selectivity of a single compound or a limited set of compounds to a range of receptors (profiles) or multiple compounds on a single receptor target. It is about analysis (advanced processing screening project). In each case, the width of the dataset is not large enough to handle the statistical correlation between the receptor targets and the chemical structures. Second, and importantly, these partial datasets are generated for compounds selected based on structural novelty, ie, potential as a new drug. Because these are new compounds, there is no biological information about their activity in the animal or human body. Thus, such an approach suffers from the same limitations as pharmacologists who attempt to interpret profile data empirically, as described above.
[0012]
[Problems to be solved by the invention]
Accordingly, it is an object of the present invention to meet the above needs by providing systems and methods for data analysis related to drug discovery and development. A full classification screening database is provided that includes positive and negative data from test results for multiple compounds and multiple molecular targets. The number of combinations of compounds and molecular targets can be determined by one of ordinary skill in the field of statistical techniques or other data mining methods by using this screening database and the associated compound and molecular target databases. It must be large enough to make reliable predictions about whether a compound is suitable for clinical trials and has a high likelihood of being a safe and effective drug.
[0013]
[Means for Solving the Problems]
The present invention discloses, inter alia, systems and methods that meet the above needs. The system includes a first database having records relating to the plurality of compounds and records relating to the effects of the plurality of compounds on human and animal biological systems, and a record relating to the plurality of molecular targets. And a second database having a second database. The computer system further includes a third database having records associated with tests for binding, reactivity, and other interactions between the compounds in the first database and the molecular targets in the second database. Including. The test includes selecting from a plurality of molecular targets in the second database and interacting with a compound known to interact with a particular molecular target (eg, a control agent or control standard). Information about the effects of a compound selected from the plurality of compounds in the first database is included, and the test is performed on a plurality of molecular targets in the second database. ing. Means for setting thresholds for interaction tests associated with said side effects, and selecting information on compounds, compound sets, and / or compounds when the results of said effect test meet said interaction test thresholds. Means are also included in the computer system. Information in the first and second databases and a record associated with one or more compounds in the first database and / or one or more molecular targets in the second database; A user interface to allow the user to view and manipulate or analyze the information in the database of Step 3, especially for compounds, molecular targets and other database records related to results meeting said interaction test threshold. Is provided.
[0014]
In addition, the invention relates to applying statistical and other data mining techniques to these multidimensional databases to determine correlations or patterns related to new drug discovery and development.
[0015]
Both the foregoing general description and the following detailed description provide examples and explanations, but are not intended to limit the scope of the invention.
[0016]
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, explain the advantages and principles of the invention.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
BRIEF DESCRIPTION OF THE DRAWINGS Preferred embodiments of the present invention will be described, examples of which are illustrated in the accompanying drawings and will be apparent from the detailed description of the invention. The same reference numbers in different drawings shall identify the same or similar elements wherever possible.
[0018]
The systems and methods consistent with the present invention analyze data related to new drug discovery and development, e.g., the potential for new compounds to be safe or effective new drugs, and It is possible to predict whether to proceed. In the following description, the system and method of the present invention will be used in relation to a relational database containing a plurality of main tables, and the binding between a compound and a molecular target will be used as a metric for the interaction between the two. Explanation will be given in relation to usage. This description may also generally apply to other database structures having multiple main components, and to measuring other interactions between compounds and molecular targets.
[0019]
The present invention relates to novel designs, structures, and applications of information-rich databases on compounds, molecular targets, particularly proteins, other macromolecules, and the biological activities of these compounds. The present invention further provides for the identification of known drugs and drug candidates that have not passed all clinical trials or clinical trials, along with preclinical and clinical data, including side effects, reaction mechanisms, and other medical data, of these compounds in a database. It also relates to the method used as a source of information for the library. The present invention also determines the binding and other interactions between compounds in the database and molecular targets, and uses relationship analysis and data mining techniques to relate these interaction patterns to new drug discovery and development. The specific chemical structure, substructure, or other characteristic of a compound that interacts with a particular biological reaction, or the biochemical characteristics or structure of a molecular target that interacts with such , And its correlation with other features. The following can be referred to as an example of such a data mining technique. All of these references are included in the present invention.
a) Chen et al. Recursive Partitioning Analysis of a Large Structure Structure-Activity Data Consulting Monthly Review of a Three-dimensional Descriptor of a Large-Scale Structure-Activity Dataset Using a Three-Dimensional Descriptor. ;
b) Hawkins et al. , Analysis of Large Structure-Activity Data Set Using Recursive Partitioning using Quantum., "Analysis of a Large Structure-Activity Data Set Using Recursive Partitioning." Struct. -Act. Relat. 16: 296-302 (1997);
c) DePriest et al. 3D-QSAR of angiotensin converting enzyme and a thermolysin inhibitor; comparison of CoMFA models based on a priori, empirically determined active site geometry based on reduced and experimentally determined active site geometries), J. Am. Am. Chem. Soc. 115: 5372-84 (1993);
d) Good et al. Lipkowitz, K., in Reviews in Computational Chemistry; B. Boyd, D .; B. (Des.), VCH, New York, Vol. 7. pp 67-117 (1996);
e) Marshall et al. Computer-Assisted Drug Design ACS Symposium Scrica 112; American Chemistry Society: Washington, DC, 1979; pp 205-226;
f) Molec et al. , 3-dimensional structure activity relationships and biological receptor mapping (A three-dimensional structure activity of relationship and biological receptor mapping), Mathematics and Computational Concepts in Chemistry; Ellis Horwood; Chichester, 1985 years; pp 225-251;
g) Mayer et al. , A unique geometry of the active site of angiotensen-converting enzyme continuance activity structure. Comput. Aided Mol. Des. , 1: 3-16. (1987);
h) Sheridan et al. , An ensemble approach to distance geometry method: application to the nicotinic pharmacophone, j. med chem. . 29: 899-906 (1986);
i) Martin et al. , A fast new approach to pharmacophone mapping and its application to dopamine and benzodiazepine antagonists (A fast new approach to pharmacophone mapping and it's application to dopaminegic edz. Comput. Aided Mol. Des. , 7; 83-102 (1993);
j) Catalyst / Hypo Tutorial, version 2.0, BioCAD Corp. Mountain View, CA, 1993;
k) Sprague, P .; W. Automated generation of chemical hypotheses and database search by catalyst (Automated chemical generation and database searching with Catalyst), Perspect. Drug Discov. Des. , 3: 1-20 (1995);
l) Barnum et al. (Identification of common functionals ammonium molecules), J. Am. Chem. Inf. Comput. Sci. , 1996, 36: 563-71 (1996);
m) Hiphop Tutorial, version 2.3; Molecular Simulation Inc. Sunnyvale, CA, 1995;
n) Davies, K .; And Uppin, R .; , 3D pharmacophore searching, net. Sci. , (Http://www.org/Science/Cheminform/feature02.html);
o) Goldener, V.A. And Vesterman, B .; , APEX 3D expert system for drug design, Net. Sci. (Http://awod.com/netsci/science/compchem/feature09.html);
p) Van Drie, J. et al. Strategies for the determination of pharmaceutical 3D database queries, J. Pharmacophore, 3D database queries. Comput. Aided Mol. Des. , 11: 39-52 (1997).
q) Van Dire, J .; And Nugent, R .; , Addressing the challenges presented by combination chemistry: 3D databases, pharmacophon; recognition and beyond, SAR QS RNAR. Res. , 9: 1-21 (1998);
r) Finn et al. , The discovery of Pharmacophonre using Inductive Logic Programming progol (Pharmacophonre discovery using the inductive logic programming progol) Machine Learning, Special Issue on Applications and Knowledge Discovery, Kluwer Academic Publishers: Boston, 1998, pp 1-33;
s) Jain et al. J., a shape-based machine learning tool for drug design, Comp. Comput. Aided Mol. Des. , 8: 635-52 (1994).
[0020]
Unlike standard operating procedures in the pharmaceutical industry, Group 4 databases must be constructed as databases consisting of more than two components and cover a considerable range in both receptors or enzyme targets and compounds. The background section suggests it must be. As an example, to build a three-component database, one first selects a broad set of compounds that contains a wealth of information directly relevant to drug discovery and development. The most relevant information is often obtained from the actual experience of testing such compounds in humans in clinical trials and / or post-marketing surveillance and in animals in preclinical trials. Other relevant biological information can be obtained from natural substances that exhibit one or more biological activities, or from chemical reference standards used in the art to study the biological characteristics of receptors. Thus, one embodiment of an information-rich compound selected for such a Group 4 database includes over-the-counter drugs, drugs that failed clinical or preclinical studies, natural bioactive substances or natural Extracts and control agents used in receptor binding assays.
[0021]
Such a database can be constructed using screening data obtained from the scientific literature. This approach could produce a partial dataset, but has limitations. First, literature references provide only one piece of positive information (eg, a report on the inhibition of binding of a particular compound to a particular receptor). Negative data is as important as positive data in order to compare useful information. In addition, certain statistical analyzes may not be applicable to datasets that do not have both positive and negative data. Second, a separate quantitative report of an article on the binding data of one compound to one receptor and a separate report on the binding data of another compound for the same receptor describe how the assay is performed. Because they are different, they cannot be compared. Thus, one embodiment for generating a Group 4 three-component database is to screen a wide range of compound arrays for a wide range of receptors or enzyme targets to obtain consistent comparisons and to obtain positive and negative results. It will secure both data.
[0022]
Compound components: selection of compound libraries and inclusion of compound data
The present invention relates to a database containing, as one component, compounds having known biological activities related to pharmaceutical research and development. Information about biological activity can be included in a database or table of compounds.
[0023]
For example, these information-rich compounds include:
(A) A pharmacological control or reference compound for measuring the interaction or molecular binding between an unknown compound and a particular molecular target such as a receptor or enzyme. Examples of such control compounds include compounds used to characterize the binding effect between a test compound and a molecular target, including a receptor or an enzyme. Other control agents include Sigma Aldrich Corp. Research Biochemicals Inc., a group of Compounds selected from catalogs of (RBI) and other sources known in the art may be included. Because these pharmacological control compounds are pre-tested and / or marketed as medicaments or are highly bioactive natural substances, they may overlap with the following three categories:
(B) Compounds that are known drugs that are currently or previously marketed for medical use and for which substantial amounts of biological information are available. These compounds are well known and appear in publications available from U.S. government agencies such as the Federal Food and Drug Administration (FDA), as well as in publications issued by private companies or nonprofits. One such publication, issued by a non-profit organization, is the USP DI Series by the United States Pharmacopeial Convention Inc., in which Chapter I. Drug Information for Healthcare Providers (Volume I. Drug Information for the Health Care Professional) is updated monthly by the USP DI Update. New drugs that have been licensed will fall into this category. Over-the-counter drugs or drugs approved by the FDA or equivalent foreign regulatory bodies are public records, so those of ordinary skill in the art will readily identify such compounds as belonging to this category. I can do it.
(C) Compounds given IND (Investigational New Drug) status as potential new drugs, but failing to achieve sufficient efficacy or safety in clinical trials to obtain FDA approval; Compounds that have been approved for human testing, such as compounds that did not achieve commercial status. This category of compounds may also include those compounds that have been approved for marketing by the FDA but have subsequently withdrawn from the market. These compounds also contain significant amounts of useful biological information and will be particularly useful for the purposes of the present invention. The identity of the failed drug can be obtained from a number of sources, including official publications from pharmaceutical or biotechnology companies, publications such as "Pink Sheets", and lists maintained by the FDA.
(D) Compounds exhibiting biological activity obtained from natural resources such as plants, microorganisms and animals. These natural substances may include toxins, antibacterials, behavioral modifiers, defenses, and other categories of compounds that provide information relevant to new drug discovery and development. Identification of natural substances can be obtained from many publications such as, but not limited to, RBI and Sigma Aldrich's Compound Catalog.
[0024]
For each of the compounds included in this database, the chemical structure, chemical formula, physicochemical properties, chemical spatial configuration, other spatial chemistry information (eg, Smiles codes), solubility, and other relevant data are available. In the database field. Those of ordinary skill will be able to recognize other recordable parameters. Compounds can be constructed from relationships of chemical structures in the database or from other relationships.
[0025]
FIG. 1A shows a compound table 300 of a relational database. The table 300 lists a plurality of compounds and includes a record of a plurality of compounds N (row 1 行 N). For each compound, there is a column 301-307 containing information about the compound. For example, in FIG. 1A, column 301 contains the name of the compound, column 302 contains the type of the compound (eg, a compound permitted to be tested in the human body), and column 303 contains information about the chemical structure, for example, a structural diagram. A hyperlink to invoke the screen (see snapshot 310 in FIG. 1B), column 304 shows the chemical formula of the compound, column 305 shows information on the physicochemical properties of the compound, column 306 shows the spatial arrangement of the compound, and column 307 contains information on the solubility of the compound.
[0026]
Additional columns may be added so that other relevant data for each compound 301 listed in table 300 can be included. By including the biological activity of the compounds in these added columns, the compound database can also be a two-component database (see database 500).
[0027]
FIG. 1B shows a snapshot 310 that includes information related to the records in table 300. For example, the chemical formula 304 of the compound can be included in the snapshot along with the structure 303 of the compound.
[0028]
Molecular target components: selection of receptors, enzymes, and other molecular targets and inclusion of molecular target data
The second component of the database of the present invention includes, as molecular targets, receptors, enzymes, other proteins, nucleic acids, carbohydrates, other high molecular compounds, etc. related to drug discovery and development. In one embodiment of the invention, receptors and enzymes are the primary molecular targets. Receptors alleviate most of the molecular communication in cells and organs in the body. Enzymes often enhance such communication, for example, through secondary messenger systems and cellular signaling pathways.
[0029]
Receptors include typical receptors such as dopamine receptors, serotonin receptors, opiate receptors, muscarinic receptors, adrenergic receptors, adenosine receptors, and the like. These receptor groups include receptor type subtypes (such as dopamine-1, dopamine-2, dopamine-3, dopamine-4 and dopamine-5 receptors). Certain subtypes have additional variations (such as dopamine 4.2, dopamine 4.4 and dopamine 4.7), or some have different shapes (such as dopamine 2 short and dopamine 2 long). Mutations in the gene encoding a particular receptor may lead to a subset of receptors that have a slightly different binding to drugs and other compounds than normal receptors, resulting in splice variants of the receptor There is a possibility. Receptors can be categorized by family, superfamily, or subfamily. There are classification methods such as G-protein binding receptor, transmembrane receptor, and nuclear receptor. Receptors can be classified according to the degree of similarity of the DNA sequences of the related genes. Receptors can also be classified by amino acid sequence and related three-dimensional coordination. Receptors can be classified by their location in the tissue or across different cell types.
[0030]
Enzymes include proteases, carbohydrases, kinases, phosphotases, DNA modifying enzymes, transferases, P450s, and other enzymes known to those skilled in the art.
[0031]
Applicants are constantly developing other receptors, receptor sources, and assays related thereto and adding them to the database content. Additional receptors and receptor assays are well known to those skilled in the art. Lists and descriptions of receptors relevant to drug discovery and development can be obtained from many publications known to those skilled in the art. These publications include the RBI Handbook of Receptor Classification and the IUPHA Receptor Classification Book. Further, as new receptors and receptor subtypes are discovered, they are also added to the database content.
[0032]
Enzymes and enzyme assays are well known to those skilled in the art. Lists and descriptions of receptors relevant to drug discovery and development can be obtained from many publications known to those skilled in the art.
[0033]
FIG. 2 shows tables 400, 410 and 420 that form part of a relational database system that can be used to access molecular target information. Table 400 lists the targets and includes a record of multiple targets M (lines 1-M). Column 401 lists the target names, and column 402 specifies the type of target for each target name.
[0034]
The table structure may vary depending on the type of target identified in column 402. Table 410 contains information about targets that are classified as receptors and listed in table 400. By querying the database for a particular receptor name, the records in table 410 can be accessed. The receptor names in table 410 can be accessed by querying table 400 for a target name indicated as “Receptor” in column 402.
[0035]
The column 411 of the table 410 contains the receptor name, and this receptor name is also the target name of the column 401 of the table 400. Column 412 contains receptor family information, column 413 contains receptor superfamily information, column 414 contains receptor subfamily information, column 415 contains information about the degree of similarity of the DNA sequences of related genes, Column 416 contains information about the amino acid sequence. Amino acid sequences are one of many molecular descriptors contained in databases. Other molecular descriptors 417 may include, for example, hydrotherapy plots related to amino acid sequences. Since the molecular target database shown in Tables 400, 410 and 420 contains target information, and the biological information associated with the target is also included in the database (Table 600), the database has two components. Conceivable. The columns shown here indicate the types of information that can be included in the database and should not be construed as limiting the invention.
[0036]
The table 420 includes information on targets classified as enzymes in the table 400. By querying the database for a particular enzyme name, the records in table 420 can be accessed. The enzyme names in table 420 can be accessed by querying table 400 for the target name indicated in column 402 as "Enzyme".
[0037]
The enzyme name is included in the column 421 of the table 420, and the enzyme name is also included in the target name in the column 401 of the table 400. Column 422 contains information about the type of enzyme. The column 423 is set as “other related information”, and an additional column may be added to the table 420 depending on, for example, a case where a user wants to access other enzyme information including an amino acid sequence and a molecular descriptor. Is shown.
[0038]
Although only tables 410 and 420 are shown to illustrate access to molecular target information by type of target, the relational database system provides additional information depending on the number of types of molecular target targets available in the database. Tables can be added.
[0039]
Biological Information Component: Biological / Chemical Information Parameters
Biological information that forms part of the database includes, for example, items related to side effects, mechanisms of pharmacological actions, metabolism by drugs, toxicity, absorption, distribution, and elimination. This information can be obtained from FDA-approved labels for over-the-counter drugs or from the literature and publications on drugs that failed clinical trials. Specific examples of parameters include toxicity, LD₅₀, LD₅₀/ ED₅₀Teratogenicity, toxic mechanism, toxic target organ, in vitro toxic battery, apoptosis inducing, bioavailability, absorption, blood-brain barrier, oral absorption, mucosal absorption,% absorption , Distribution, critical blood protein, half-life, onset of action, duration of action, peak in blood concentration, metabolism, major pathway, minor pathway, active metabolite, excretion, first elimination mode, second elimination mode, These include in vitro utilities, therapeutic indications, effects on animal behavior, side effects, known key targets, other target organs / systems, and known receptor interactions.
[0040]
FIG. 3 shows a table 500 containing some of the biological information parameters described above. Table 500 consists of N rows (1-N) relating to all compounds that may be present in the first database. Column 501 contains the compound name, column 502 contains treatment instructions (for over-the-counter drugs or drugs that did not pass the test), column 503 contains information about toxicity, and column 504 contains information about toxicity. Contains information about side effects, and column 505 contains information about the mechanism of action of the drug. For example, associating table 500 with table 300 can form a two-component table of compounds and biological activity.
[0041]
FIG. 3 also shows a table 600 containing biological information parameters associated with molecular targets in the database. The table 600 relating to all possible targets in the second database consists of P rows (1 to P). Column 601 contains the target name, column 602 contains the treatment instructions (for over-the-counter drugs or drugs that did not pass the test), column 603 contains information about toxicity, and column 604 contains Contains information about side effects. Similarly to the above, for example, by associating the table 600 with the table 400, a two-component table of a molecular target and a biological activity can be formed. Tables 500 and 600 together provide a complete taxonomy database (eg, all combinations of compounds and molecular targets that may exist in a relational database system) including molecular target information, compound information, and biological activity information associated with each molecular target and each compound. And can be regarded as a multidimensional database. Additional columns can be added to tables 500 and 600 without departing from the scope of the invention.
[0042]
Determination of connectivity information
A key feature of the present invention is the composition of multiple information components consisting of compounds, molecular targets, and biological information, and the evaluation of the binding, reactivity, and other interactions between the compounds and molecular targets. By reassociating this binding or reactivity information with known biological information, patterns or relationships that can be used for drug discovery and development can be screened. An important aspect of the present invention is to generate as extensive and consistent binding or reactivity data between compounds and molecular targets as possible to provide as complete a dataset as possible to identify the required patterns or relationships. And provide both positive and negative binding or reactivity information. In one embodiment of the present invention, for example, the binding data is configured as a numerical descriptor indicating whether or not a threshold set is satisfied for a specific molecular target or molecular target set. This numerical descriptor can also be associated with the presence or absence of reactivity for each compound and each receptor, and other molecular targets, evaluated at a concentration near a threshold suitable for the biological system or set of biological information. For example, for some compounds, the inhibition of binding between the receptor and the corresponding^-5M (10 micromolar), can be tested at a threshold of 30%. In addition, an initial concentration or an inhibition rate threshold can be set. Also, in one embodiment of the present invention, compounds that cause binding inhibition above the threshold in this first yes / no test are further tested for their ability to inhibit binding. For these active compounds, for example, 10^-5-10^-9The test was performed in a series of concentration conditions, including 7-14 different concentrations within the range of M, and the IC of the active compound at a particular receptor was tested.₅₀And / or determine the value of Ki. The number of concentration conditions for making this determination is at least better, at least^-5-10^-9It may be necessary to use a concentration range higher or lower than M. From these data, a matrix of relative activity or relative potency of each active compound for each molecular target is obtained.
[0043]
To generate these screening data, the compounds are first solubilized in a suitable solvent system. As the solvent system, for example, 4% DMSO or the like can be used, but other concentrations of DMSO or another solvent can also be used. The stock solutions of these compounds are then diluted to an appropriate concentration to make them available as a repository. Each assay for measuring the interaction between a compound and a molecular target will use different reagents and procedures. Each such assay needs to be characterized and routinely made consistent. An appropriate control test must be performed each time the assay is run. Any assay format that can produce information of the type and accuracy desired can be used. Numerous assay detection systems can be used, such as radiolabeling, fluorescence analysis, fluorescence polarization analysis, time-resolved fluorescence analysis, fluorescence correlation spectroscopy, chemiluminescence, UV absorption, colorimetry, and the like.
[0044]
In one embodiment of the invention, data regarding molecular interactions is generated using a receptor binding assay or an enzymatic activity assay. For example, in a receptor binding assay, compounds in a stock solution are tested for their ability to inhibit the binding interaction between the receptor and a reference agent selected for the receptor. The receptor can be obtained from animal or human tissues or the like, or can be obtained from a cell line transfected to contain the gene for the receptor. The receptor source for the assay can be provided, for example, as a cell fraction containing the receptor. The receptor may also be partially purified. A reference compound or ligand is preferably selected based on potential and / or specific binding to a particular receptor, and by including iodine 125, tritium, carbon-14, and other radioactive tracers, Bound and unbound ligands can be distinguished. Positive and negative control tests are performed in parallel with the testing of the binding data of the compounds to be included in the database, and reference curves at various concentrations of the reference (radioactive) ligand ensure the quality of the assay performed. .
[0045]
One of skill in the art will appreciate that numerous methods and systems allow the interaction between a target and a compound to be measured. To reach an equilibrium state in the binding reaction, the radioligand, the receptor preparation solution, and the test compound are cultured for an appropriate time in an appropriate buffer solution and at an appropriate temperature. The amount of bound radioligand relative to unbound radioligand is determined by filtration and separation steps using methods such as SPA (scintillation proximity assay) and is measured by liquid scintillation or gamma counting. The specific binding number of the test compound is then determined by comparing the test compound assay results with the positive and negative control tests. From these data, the percent inhibition of the test compound is calculated.
[0046]
FIG. 4 shows a table 200 representing a screening result and an assay database, wherein compounds contained in database 300 (comprising 1 to N compounds) are tested for action on molecular targets contained in database 400. It has become. Table 200 can take many forms. For example, in table 210, the screening results based on whether the screening result of each of the plurality of compounds tested against each of the plurality of molecular targets is above or below a test result threshold selected for each measurement set. Can be entered as a “yes” or “no” entry.
[0047]
In other examples, the screening results identify the potency or degree of binding or other effect (eg, the Ki of the compound-receptor interaction) for each of the plurality of compounds tested against each of the plurality of molecular targets. It is input to the table 220 as a numerical descriptor. In a preferred embodiment, all such "compound" x "target" matrix points are determined in tables 210 and 220, and an all-kind database is generated. In addition, the screening result and assay database 200 may include measurement results of other compound-target interactions, including the raw data of the screening results, the measurement results obtained from the raw data, and the assay results. Procedures and features, and other relevant information are included.
[0048]
FIGS. 5A and 5B illustrate the use of the database 100, illustrated here as receptor selectivity, as part of the screening process to discover and screen new compounds as new drug candidates to be further developed. FIG. 5A, or finding and selecting new targets as potential effective targets for use in finding new drug candidates for specific disease detection (FIG. 5B). The database 100 may include a compound component 300, a molecular target component 400, biological information components 500 and 600, and a screening results and assay database 200.
[0049]
New compounds or sets of compounds are introduced into the screening process 102 to determine whether they have the ability to inhibit binding to specific compounds (eg, reference agents) and molecular targets (see FIG. 5A). In the screening process, target information obtained from the molecular target component 400 can be used.
[0050]
The results of the screening process 102 can be stored in an intermediate database or entered into the screening results and assay database 200 of the receptor selectivity database 100. In addition, the result can be stored as a specific parameter (for example, cytotoxicity) in the biological information database 500 or stored in the compound database 300 (for example, as a compound name).
[0051]
The complete set of results from the screening process 102 can be stored in the screening results and assay database 200. These new compounds can be further tested by querying the database 200 for new compounds (eg, reference agents) that exhibit an inhibitory capacity for binding between the compound and the molecular target.
[0052]
Alternatively, new molecular targets, such as, for example, "orphan" receptors, can be introduced into the screening process to test for compounds in the compound database 300 (see FIG. 5B). Regarding the orphan receptor, its structure is known, but its function and its relation to disease are not known. The results of the screening process, including the identification of compounds that interact with the new molecular target, are incorporated into the screening results database 200. The database 100 is queried to identify the function of the new molecular target and / or to confirm the relevance of the new molecular target to the disease.
[0053]
FIG. 6A shows the prediction of the potential of a new compound as a drug using the database 100. Table 710 relies on information from a database of compounds (300), molecular targets (400), biological information (500 and 600), and screening results (200). When a user provides information about a new compound to the database 100, the automatic query script is executed to obtain this information so that the table 710 includes information from one or more of these databases (or tables). It has become.
[0054]
A query script for generating the table 710 can select a compound from the compound database 300 when given information on a new compound. This selection can be made based on the similarity in chemical structure and other properties between the new compound and the compounds already included in database 300.
[0055]
After selecting the compound, the query script selects from the target database 400 targets that are known to react (bind) with the selected compound. Finally, the biological information databases 500 and 600 are queried using the selected compound and molecular target, and the biological information for the compound-molecular target pair is inserted into the table 710. Alternatively, the biological information contained in the table 710 can be limited to a particular biological information category of interest (e.g., toxicity) by the user entering the category.
[0056]
The user can query table 710 to obtain information related to predicting the potential of a new compound to be used as a drug. Examples include queries for molecular targets that are known to react with compounds related to the novel compound, and queries for side effects that the molecular targets are known to cause with the compound.
[0057]
FIG. 6B shows the association of novel molecular targets with diseases and / or organisms using the data entry and query shown in FIG. 6B, using an approach similar to predicting the drug potential of new compounds using database 100. The method for confirming the biological function is shown.
[0058]
All patents, patent applications and publications mentioned herein are hereby incorporated by reference.
[0059]
The above description of the embodiments of the present invention is for the purpose of illustration and description, is not exhaustive, and does not limit the invention to the forms disclosed herein. Modifications and changes may be made to the invention in light of the above disclosure or upon practicing the invention.
[Brief description of the drawings]
FIG.
FIG. 1A shows a compound table of a receptor selectivity mapping database in one embodiment of the present invention.
FIG. 1B shows a snapshot of a compound record including the spatial arrangement of compounds in a receptor selectivity mapping database in one embodiment of the present invention.
FIG. 2
FIG. 4 illustrates several logical tables of a receptor selectivity mapping database that can be used to access molecular target information in one embodiment of the invention.
FIG. 3
4 shows a table of biological information in a receptor selectivity mapping database in one embodiment of the present invention.
FIG. 4
FIG. 4 illustrates a method of using a receptor selectivity mapping database as part of a screening process in one embodiment of the present invention.
FIG. 5
FIG. 5A shows how to use the receptor selectivity mapping database as part of the screening process to find and select new compounds that are new drug candidates.
FIG. 5B shows how a receptor selectivity mapping database is used as part of the screening process to identify new targets that are candidate targets for finding new drug candidates for a particular disease.
FIG. 6
FIG. 6A shows the use of a database to predict the potential of a new compound as a drug.
FIG. 6B illustrates the use of a database to identify disease relevance and / or biological function of novel molecular targets.

Claims (26)

A first database comprising records relating to the plurality of compounds and records relating to biological information regarding the effects of the compounds on the biological system;
A second database containing records relating to the plurality of molecular targets;
A test of an interaction between a compound of a first database and a molecular target of a second database, wherein the interaction between the compound and the molecular target known to interact with the plurality of molecular targets is performed. A third database comprising a record relating to the test for the action, the information including information of the action of one of the plurality of compounds;
A user browses the selected compound and, depending on its association with the record of compound data in the first database or the record of molecular targets in the second database, the first database, the second database and the third database. A user interface that can selectively browse information from the database of
Computer system consisting of The computer system according to claim 1, wherein the interaction includes binding, and the effect includes an inhibitory effect. The computer system according to claim 1, wherein the compound includes a compound having no known biological activity or a compound that does not pass a test. The computer system of claim 1, wherein the compound comprises a compound that has been tested in an animal. The computer system according to claim 1, wherein the compound includes a compound known to act on the environment. The computer system according to claim 1, wherein the compound includes a pharmacological reference agent. The computer system of claim 1, wherein the compound includes a drug known in the clinical drug market and having a significant amount of available biological information. The computer system of claim 1, wherein the compounds include compounds that have been approved for human testing. The computer system according to claim 1, wherein the compound includes a compound exhibiting biological activity obtained from a natural resource. The computer system according to claim 1, wherein the molecular target includes a receptor. The computer system according to claim 1, wherein the molecular target includes an enzyme. The computer system according to claim 1, wherein the molecular target includes a nucleic acid. The computer system according to claim 1, wherein the molecular target includes a carbohydrate. 2. The computer system of claim 1, wherein the first database records associated with a plurality of compounds are organized by categories relating to the description and properties of the compounds. 15. The computer system of claim 14, wherein the categories include compound name, compound type, physicochemical properties, descriptor of chemical spatial arrangement or chemical structure, and solubility. The computer system according to claim 1, wherein the first database includes a database of natural substances. The computer system according to claim 1, wherein the first database includes a database of drugs that did not pass the test. The computer system according to claim 1, wherein the first database includes a chemical registry database. The computer system according to claim 1, wherein the second database includes a database having a three-dimensional structure. The computer system according to claim 1, wherein the second database includes a sequence / mutation database. The computer system according to claim 1, wherein the second database includes a gene database. The third database record relating to biological information regarding the effects of the compound on the biological target is systematically organized by category including compound name, target name, toxicity, side effects, and mechanism of pharmacological action. The computer system according to claim 1, wherein the computer system is configured. Further comprising: means for setting an interaction test threshold associated with the effect; and means for selecting the compound if the use of the compound results in meeting the interaction test threshold. 2. The computer system according to claim 1. Selecting a compound from a first database containing records related to the plurality of compounds;
Selecting a molecular target from a second database containing records relating to the plurality of molecular targets;
Generating information related to an interaction between each of the selected compounds and each of the selected molecular targets;
Selecting a biological activity from a third database containing records relating to biological information regarding the effect of the compound on the biological target;
Using the generated information to associate a pattern of interaction between a compound and a molecular target with the selected biological activity. The step of generating the information includes:
Generating binding data for the binding between each of the selected compounds and each of the selected molecular targets by monitoring the inhibitory effect of the unknown compound on the binding;
Setting a binding test threshold for the inhibitory effect,
Generating information about combinations of unknown compounds, molecular targets, and compounds that meet or do not meet the binding test thresholds. The method of claim 25, wherein the connectivity data includes positive and negative connectivity information.

Images