PolySearch2

PolySearch2 extracts and ranks associations among biomedical entities by text-mining literature and integrating multiple biological databases to support discovery of relationships between human diseases, genes, single nucleotide polymorphisms (SNPs), proteins, drugs, metabolites, toxins, metabolic pathways, organs, tissues, subcellular organelles, positive and negative health effects, drug actions, Gene Ontology terms, MeSH terms, ICD-10 medical codes, biological taxonomies, and chemical taxonomies.

Key Features:

• Supported entity types: Identifies associations among human diseases, genes, single nucleotide polymorphisms (SNPs), proteins, drugs, metabolites, toxins, metabolic pathways, organs, tissues, subcellular organelles, positive and negative health effects, drug actions, Gene Ontology terms, MeSH terms, ICD-10 medical codes, biological taxonomies, and chemical taxonomies.
• Generalized query format: Supports the query paradigm "Given X, find all associated Ys" where X and Y can be any supported biomedical entities.
• Literature corpora searched: Performs searches over local versions of MEDLINE abstracts, PubMed Central full-text articles, Wikipedia full-text articles, and US Patent application abstracts.
• Integrated databases: Incorporates records from 14 text-rich biological databases, including UniProt, DrugBank, and the Human Metabolome Database.
• Thesaurus of biological terms: Uses an extensive thesaurus to map synonyms and related terms across datasets.
• Search engine technology: Leverages advanced search-engine technology to retrieve pertinent articles and database records.
• Association ranking: Generates associative candidates and ranks them based on relevancy statistics.
• Evidence annotation: Annotates results with highlighted key sentences extracted from source texts.

Scientific Applications:

• Exploratory association discovery: Identify literature- and database-supported associations among diverse biomedical entities for hypothesis generation.
• Toxicant–disease mapping: Map diseases associated with environmental chemicals or toxins, exemplified by queries such as "Find all diseases associated with Bisphenol A."
• Integrative evidence aggregation: Combine evidence from MEDLINE, PubMed Central, Wikipedia, US Patents and 14 biological databases (including UniProt, DrugBank, Human Metabolome Database) to corroborate entity relationships.

Methodology:

Performs comprehensive searches of local MEDLINE abstracts, PubMed Central full-text articles, Wikipedia full-text articles, and US Patent application abstracts; integrates records from 14 text-rich biological databases (including UniProt, DrugBank, Human Metabolome Database); applies an extensive thesaurus to identify candidate associations, ranks candidates using relevancy statistics, and annotates results with highlighted key sentences using advanced search-engine technology.