Tool catalog

Download one or more public datasets into the sandbox, then run a Python analysis script over them. Fetches via curl with structured failure classification (paywall / dns / timeout / etc.) so the agent can ask the user for a workaround. Chart outputs should follow Cortexa's standard scientific style and save both figures and chart-ready data under artifacts/.

Inspect the schema of one SQL connection the caller owns (Postgres or BigQuery): tables and their columns, restricted to the schemas or datasets that connection allows. Reads structure, not rows.

Reads a data source you connect — not public data.

Apply typed cell-level edits to an owned .xlsx workbook without regenerating it. Resolves the exact Library/session/analysis source server-side, applies set_cell or clear_cell operations with openpyxl, reloads the serialized result, rejects scope drift, and returns a versioned workbook plus a machine-verifiable cell diff.

Get commit metadata for a granted private repository and ref: the resolved immutable SHA, author, message, and GitHub URL. Use it to pin what a citation of the caller's code actually refers to.

Reads a data source you connect — not public data.

List the private repositories granted to a GitHub connection the caller owns, via the Cortexa GitHub App. Returns immutable numeric repository ids to use with the tree, search, and read tools. The app holds no write permissions.

Reads a data source you connect — not public data.

List source paths in a granted private repository at an immutable commit, so the agent can learn a project's structure before reading files. Bounded by the connection's allowed path prefixes.

Reads a data source you connect — not public data.

Read a bounded text range from a granted private repository. Resolves the ref to a commit SHA and returns an immutable permalink, blob SHA, exact line range, and text, so an answer can cite code that will not move.

Reads a data source you connect — not public data.

Search code paths across the private repositories granted to a GitHub connection the caller owns. Search identifies candidate paths; read the file to ground an answer in commit-pinned content, since search URLs follow the mutable default branch.

Reads a data source you connect — not public data.

List the private data connections the signed-in caller is allowed to read — the ones they configured in Settings → Connections. Returns each connection's id, provider, name, allowed schemas, and last test status, never any of its contents. Returns an empty list when the caller has configured none.

Reads a data source you connect — not public data.

Run one read-only SQL statement against a Postgres or BigQuery connection the caller owns and return the rows as research data. Writes and DDL are rejected; Postgres runs inside a read-only transaction scoped to the connection's allowed schemas, and BigQuery enforces a per-query scan-byte ceiling. Results are row-capped and flag truncation.

Reads a data source you connect — not public data.

Render an inline chart (bar, line, scatter). Capped at 30 data points by design — the agent should summarize, not table-dump.

Elevate a single citation into an inline card. Best for the 1–3 most important sources backing an answer; less-important sources stay as inline [source:id] markers.

Render a LaTeX formula inline. Use for binding affinity, dose-response, or any key math that helps the answer.

Render a molecular structure inline (PDB id, AlphaFold UniProt id, or remote mmCIF/PDB URL). Supports residue highlights with labels and colors.

Execute Python in an isolated Cortexa compute sandbox (Python 3.13). Default packages: numpy, pandas, scipy, matplotlib, scikit-learn, statsmodels, requests, seaborn, plotly, kaleido, pyarrow, openpyxl, beautifulsoup4, lxml, pillow, duckdb. Chart outputs should follow Cortexa's standard scientific style and save both figures and chart-ready data under artifacts/.

Search arXiv for preprints across CS/ML, physics, math, quant-bio, and more. Use for methodology papers, bleeding-edge algorithms, and pre-peer-review results.

bioRxiv + medRxiv preprints. Use for the most recent biology and clinical research (pre-peer-review).

US Bureau of Labor Statistics: authoritative US labor & price series — CPI inflation, unemployment rate, nonfarm payrolls, average hourly earnings, PPI, and labor-force participation. Use for the official US jobs/inflation release figures. Returns the latest datapoint with its release period.

Search ClinicalTrials.gov. Use for trial phase, status, enrollment, primary endpoints, and sponsor. Returns NCT records.

Search ClinVar for clinical significance of genetic variants. Use for variant pathogenicity, allele frequency, and condition associations.

CoinGecko: live cryptocurrency market data — price, market cap, 24h volume, rank, supply, and all-time high across 10K+ coins. FIRST CHOICE for any crypto/digital-asset price or market question. Search by coin name or symbol (e.g. 'bitcoin', 'ETH', 'solana').

Congress.gov: bills, resolutions, members, roll-call votes, and committee reports from the Library of Congress. Use to track the status of legislation, find sponsors, or look up Congressional voting records.

CourtListener (Free Law Project): US federal + state case law (9M+ opinions, including the historical Caselaw Access Project corpus) and RECAP federal dockets. FIRST CHOICE for any question about a court opinion, judicial decision, or active litigation, AND for historical precedent (Standard Oil, Brown v. Board, Sherman Act cases). Uses semantic + hybrid search by default — write the query in natural language ('cases where shareholders sued directors for ignoring climate risk') and put any term that MUST appear literally in "double quotes" (e.g. '"Section 230" immunity for AI-generated content'). Returns case name, court, citation, date filed, judge, and excerpt.

Crossref: canonical DOI registry with bibliographic metadata (journal, ISSN, licence, publisher). Use when the user gives you a DOI to resolve, or when you need authoritative journal/book metadata. Also useful as a backup when OpenAlex is rate-limiting.

DailyMed NIH prescribing information. Use for the official FDA-cleared package insert text.

Search the Directory of Open Access Books (DOAB) for peer-reviewed OA scholarly books across disciplines. Use for broad open-access monographs and edited volumes when you need a book-length OA source outside NCBI Bookshelf or OpenStax. Returns title, abstract when available, and DOAB landing URL.

ECB Data Portal: euro-area headline indicators — reference exchange rates (EUR/USD, EUR/GBP, EUR/JPY), ECB policy rates (main refinancing, deposit facility), and HICP inflation. Use for euro-area monetary policy, FX, and inflation questions. Returns the latest official value with its period.

Electronic Code of Federal Regulations (eCFR): authoritative current text of all 50 CFR titles. Use for any question about federal regulations — securities (17 CFR), drugs and food (21 CFR), environment (40 CFR), labor (29 CFR), tax (26 CFR), etc. Pass an explicit citation ('21 CFR 201.25', '16 CFR 1700.14', '17 CFR 240.10b-5') to get that section's authoritative current text plus any pending amendment; pass a short topic phrase (3-6 specific terms, not a full sentence) to search by relevance. Never state a CFR section number you have not retrieved here.

SEC EDGAR full-text filings search. Use for company filings — 10-K (annual), 10-Q (quarterly), 8-K (material events), S-1 (IPO), proxy statements, Form 4 (insider trades). FIRST CHOICE for questions about a public company's risk factors, financials, governance, or insider activity.

EUR-Lex: EU legal portal — treaties, regulations, directives, decisions, and Court of Justice (CJEU) case law. FIRST CHOICE for any EU-law question (GDPR, AI Act, DSA/DMA, environmental directives, competition rulings). Returns CELEX identifier, document type, date, and excerpt.

Search FDA-approved drug labels. Use for indications, contraindications, dosing ranges, adverse reactions, and black-box warnings. Returns OpenFDA label records.

FRED (Federal Reserve Economic Data): 800K+ US & international economic time series — interest rates (DGS10, FEDFUNDS), inflation (CPIAUCSL), GDP, unemployment (UNRATE), payrolls, money supply (M2), exchange rates, and equity indices (SP500). FIRST CHOICE for any macroeconomic, monetary-policy, or markets time-series question. Returns the matched series with its latest value and as-of date. (Requires FRED_API_KEY.)

GitHub: search public repositories by name, description, topics. Use to find official implementations of papers, kernel / compiler / database projects, RFC drafts, and to gauge reproducibility / maintenance for any CS/ML work.

GovInfo (GPO): Federal Register, public laws, US Code, Congressional reports, Compilation of Presidential Documents. Use for proposed rules, agency notices, statutes, and Congressional hearings. Complements eCFR (final regulations) by covering the rulemaking process.

Human Phenotype Ontology. Use to map symptoms to phenotype terms and associated conditions.

Hugging Face Hub: ML models + datasets registry. Use for anything about pretrained models, benchmark datasets, model cards, licences, and downloads/likes signal. First-choice source for CS/ML reproducibility queries.

ICD-10/11 diagnosis coding. Use to map a clinical description to a standardized diagnosis code.

LOINC lab and observation codes. Use when you need the LOINC code for a specific lab test or measurement.

NASA ADS (Astrophysics Data System): canonical literature index for astrophysics, cosmology, planetary science, and space sciences. Use for any astronomy/astrophysics query — it covers ApJ, MNRAS, A&A, IAU, arXiv astro-ph, and more.

Search NCBI Bookshelf for biomedical and clinical textbooks, chapters, and long-form reference (NBK accessions). Use for textbook-style clinical background, disease overviews, and NCBI book chapters — not for journal articles (use PubMed) or point-of-care CDS. Returns titles, book context, and Bookshelf URLs; free-to-read snippets only (not full-text scrape).

Look up gene records (RefSeq, genomic context, orthologs, expression). Use when a gene symbol needs authoritative NCBI data.

OMIM catalog of Mendelian disorders. Use for gene-disease relationships and inheritance patterns.

OpenAlex: the universal scholarly-works catalog. 260M+ papers across every discipline with citation counts, open-access metadata, and concept tags. Use as the FIRST-CHOICE literature source for any question that isn't obviously biomedical, especially CS/ML, physics, chemistry, earth sciences, social sciences, economics, or humanities.

openFDA Animal & Veterinary adverse events (FDA Center for Veterinary Medicine): adverse-event reports for animal drugs — species/breed treated, drug and route, VeDDRA-coded reactions, and outcome. FIRST CHOICE for veterinary pharmacovigilance and animal-drug safety questions. Search by active ingredient (e.g. 'carprofen'), reaction, or species.

Search OpenStax for foundational STEM and undergrad open textbooks (CC BY): anatomy, physiology, biology, chemistry, physics, math, economics, and more. Searches inside the books, so a concept query ('cardiac cycle', 'price elasticity of demand') returns the specific chapter section with its text and a citable openstax.org page URL; naming a book or subject ('biology') returns the book instead. Use for curriculum-level open textbooks and student-facing explainers — not for clinical guidelines or primary literature.

Query Open Targets for drug-target-disease associations, tractability, and evidence scores. Use for target identification and drug-repurposing hypotheses.

Orphanet rare-disease encyclopedia. Use for orphan disease prevalence, inheritance, and associated genes.

Patent litigation (USPTO OCE dataset): US district-court patent infringement cases — caption, case number, court, and filing/close dates. Use for patent-litigation history, party disputes, and 'has this been litigated' questions. Search by party or case name.

Patents (PatentsView via BigQuery): search US patents by title/abstract. FIRST CHOICE for patent-landscape, prior-art, FTO, and IP-strategy questions. Returns patent number, grant date, title, and abstract snippet, linked to Google Patents.

PubChem: NIH compound database with structures (SMILES, InChI), physicochemical properties (MW, XLogP, TPSA), and GHS hazard classifications. FIRST CHOICE for any chemistry question that names a molecule, compound, or drug substance.

Search PubMed for peer-reviewed biomedical literature. Use for clinical research, drug efficacy studies, disease mechanisms, and treatment outcomes. Returns abstracts + PMIDs.

Look up drug nomenclature, ingredients, brand/generic mappings, and drug-drug interactions. Use when you need canonical drug identifiers.

SEC Financials (XBRL company facts): structured financial-statement numbers for US public companies — revenue, net income, total assets, shareholders' equity, and diluted EPS — resolved by ticker or company name. Use when you need the actual reported figures (not the filing narrative — that's search_edgar). Returns the latest annual values with period-end dates.

Semantic Scholar: academic papers with citation counts and influence scores. Use when you need citation analysis or cross-discipline coverage.

Search UniProt for protein sequences, functions, domains, PTMs, and cross-references. Use when the question involves a specific protein (gene symbol or accession).

US Healthcare Providers (CMS NPPES NPI registry): resolve a US clinician or healthcare organization name to its NPI, practice city/state, and primary taxonomy (specialty). Use for 'who is this provider / what's their NPI' lookups and provider directories. Search by person name (e.g. 'Atul Gawande') or organization name.

Zenodo: CERN-hosted artifact registry with persistent DOIs. Use to locate versioned datasets, software releases, and supplementary research artifacts that a paper depends on. Especially good for reproducibility questions.

List the channels a Slack connection the caller owns is permitted to read. Returns each channel's id, name, privacy, and membership so a later read or search can name a channel. Channels outside the connection's allowed list are never returned.

Reads a data source you connect — not public data.

Read recent messages from one channel of a Slack connection the caller owns. Read-only and capped; channels outside the connection's allowed list are rejected server-side. Returns messages with author id, timestamp, and text.

Reads a data source you connect — not public data.

Search message history across a Slack workspace the caller owns. Requires a user token (xoxp-); bot-token connections return a clear note instead of results. Matches are filtered to the connection's allowed channels and come back with channel, author, text, timestamp, and permalink.

Reads a data source you connect — not public data.

List recent American College of Cardiology (ACC) clinical practice guidelines. Returns ACC-authored guidelines published in JACC and other ACC journals, sorted by date. ACC guidelines cover coronary revascularization, valvular heart disease, chest pain evaluation, cardiac catheterization, and cardiovascular risk assessment.

Fetch detailed content from a specific ADA Standards of Care section by PMID. Retrieves the abstract and, when available via PubMed Central (PMC), a full-text snippet. Use ADA_list_standards_sections first to get PMIDs for each section. All ADA Standards sections are freely available in PMC. Returns title, abstract, full text snippet, and links.

List all sections/chapters of the current ADA Standards of Medical Care in Diabetes. The ADA publishes the Standards of Care annually in Diabetes Care as a free supplement. Each section covers a specific clinical topic (e.g., glycemic goals, pharmacologic treatment, cardiovascular risk management, diabetes technology). Returns section titles, numbers, PMIDs, and PMC links. Use year parameter to access specific editions (default: 2026).

Search ADA (American Diabetes Association) guidelines and publications by topic using PubMed. Searches publications with ADA as corporate author, returning titles, abstracts, PMIDs, and PMC links. Useful for finding ADA recommendations on specific diabetes-related topics such as glycemic targets, insulin therapy, GLP-1 receptor agonists, SGLT2 inhibitors, gestational diabetes, diabetic kidney disease, or diabetes technology.

Search a curated registry of ~30 major longitudinal cohort studies relevant to aging research worldwide. Contains study metadata including design, sample size, age range, variable categories, and data access information. Use to find the right cohort for a specific research question, such as which studies measure iron intake, grip strength, or walking speed in older adults. Filters by country, study design (longitudinal/cross-sectional), minimum sample size, and specific variable availability.

Fetch the full text (or a detailed snippet) of a specific AHA (American Heart Association) or ACC (American College of Cardiology) guideline from PubMed Central using its PMID. Returns the abstract and, when the article is open-access in PMC, the full guideline text up to 6000 characters covering recommendations, evidence tables, and rationale. Use AHA_list_guidelines or ACC_list_guidelines first to obtain PMIDs. Many recent AHA/ACC guidelines are open-access in PMC.

Search AHA (American Heart Association) and ACC (American College of Cardiology) clinical practice guidelines by topic via PubMed. Covers guidelines published in Circulation, JACC, and other AHA/ACC journals. Useful for finding recommendations on hypertension, heart failure, atrial fibrillation, coronary artery disease, valvular heart disease, lipid management, cardiac arrest, and other cardiovascular conditions.

List recent American Heart Association (AHA) clinical practice guidelines. Returns AHA-authored guidelines and scientific statements published in Circulation and other AHA journals, sorted by date. Covers cardiovascular guidelines on topics including hypertension, heart failure, cardiac arrest, stroke, arrhythmias, and prevention.

Get gene expression experiment datasets from the Allen Mouse Brain Atlas for a given gene. Returns section data sets including plane of section, section thickness, and linked gene info. These datasets contain ISH (in situ hybridization) gene expression images. Example: 'Gad1' returns 22 ISH experiment datasets.

Get detailed information about a brain structure by its Allen Brain Atlas structure ID. Returns full structure record including name, hierarchy path, parent structure, color code. Example: structure_id=382 returns 'Field CA1' in hippocampus.

Get quantified per-structure gene-expression values (StructureUnionize) for one Allen Mouse Brain ISH experiment (SectionDataSet). Returns the actual numeric spatial expression aggregated over each annotated brain structure: expression_energy (intensity x density), expression_density (fraction of expressing pixels), sum_expressing_pixels, plus the structure record (acronym, name). This is the real spatial expression signal, not just experiment metadata. Get a section_data_set_id from AllenBrain…

Search for genes in the Allen Brain Atlas by gene symbol (acronym) or name. Returns gene metadata including Entrez ID, chromosome, homologene ID. Useful for finding gene IDs needed to query expression data. Example: 'Gad1' returns glutamate decarboxylase 1 records for mouse and rat.

Search for brain structures/regions in the Allen Brain Atlas ontology by acronym or name. Returns structure hierarchy, atlas coordinates, and color coding. Example: 'CA1' returns hippocampal Field CA1 across multiple atlases.

Search the Allen Cell Types Database for individual neuron specimens characterized by electrophysiology and morphology (human and mouse). Filter by donor species ('Homo Sapiens' or 'Mus musculus') and/or brain structure. Returns per-neuron records: donor species/sex/age/disease state, brain structure, transgenic line, and key electrophysiology features (average firing rate, input resistance, membrane tau, upstroke/downstroke ratio) plus whether a morphological reconstruction exists. Distinct fr…

Get alleles/variants and affected genomic models (mutant strains) for a gene via the Alliance of Genome Resources (mouse/MGI, frog/Xenbase, and other members). Returns allele/variant records with HGVS genomic locations and disease/phenotype flags, plus affected genomic models (genotypes/strains). Example: MGI:97490 (Pax6) returns variant_summary records such as rs231853465 (NC_000068.8:g.105516528T>C) and models such as AffectedGenomicModel MGI:2680573.

Get disease summary information from the Alliance of Genome Resources by Disease Ontology (DO) ID. Returns disease name, definition, synonyms, and classification. The Alliance integrates disease annotations from multiple model organism databases. Example: 'DOID:162' returns cancer with definition and synonyms like 'malignant neoplasm'.

Get genes associated with a disease across all model organisms from the Alliance of Genome Resources. Uses Disease Ontology (DO) IDs to find genes linked to a disease from human, mouse, rat, zebrafish, fly, worm, and other organisms. Includes both direct annotations and inferred associations from model organism phenotypes. Example: 'DOID:162' (cancer) returns TP53, BRCA1, and thousands more.

Get detailed gene information from the Alliance of Genome Resources, which integrates data from 7 model organism databases (SGD, FlyBase, WormBase, ZFIN, RGD, MGI, Xenbase) plus human. Returns gene symbol, name, species, synopsis, synonyms, genomic location, and cross-references. Example: 'HGNC:6081' returns insulin gene info.

Get orthologs and paralogs of a gene across model organisms via the Alliance of Genome Resources (used by MGI, Xenbase, and other members). Returns the stringent ortholog set with per-ortholog disease/expression annotation flags and prediction methods, plus the paralog set. Gene IDs use member prefixes (MGI:, HGNC:, RGD:, ZFIN:, FB:, WB:, SGD:, Xenbase:); bare WBGene/FBgn/MGI numeric IDs are auto-prefixed. Example: MGI:97490 (Pax6) returns gene_to_gene_orthology orthologs (e.g. Drosophila ey) a…

Get phenotype annotations for a gene from the Alliance of Genome Resources. Returns phenotypes observed when the gene is mutated or knocked out, aggregated from model organism databases. Supports all model organisms (yeast, fly, worm, zebrafish, mouse, rat, frog). Example: 'HGNC:6081' (INS) returns phenotypes like 'Abnormal circulating insulin concentration'.

Get molecular (physical) interactions for a gene via the Alliance of Genome Resources (mouse/MGI, frog/Xenbase, and other members). Returns interactor gene IDs/symbols/species, interaction type, and supporting evidence references (PubMed / member curie). Gene IDs use member prefixes. Example: MGI:97490 (Pax6) returns gene_molecular_interaction records with evidence references such as MGI:6822438 (PMID:22863532).

Search for genes across all model organisms in the Alliance of Genome Resources. Matches by gene symbol or synonym (type-ahead) and returns genes from human, mouse, rat, zebrafish, fly, worm, yeast, and frog databases. Use the official gene symbol rather than a descriptive name: 'INS' returns INS (human), Ins1/Ins2 (rat), ins (zebrafish), whereas 'insulin' matches GO terms/diseases, not genes.

Get the ligands, cofactors, and ions that AlphaFill transplants into a protein's AlphaFold model, by UniProt accession. AlphaFold predicts apo (ligand-free) structures; AlphaFill adds small molecules by homology to experimental PDB structures. Returns the distinct transplanted compounds (PDB chemical-component codes), how many homologous hits support each, the best local fit RMSD (lower = better), and the source PDB entries. Use to hypothesize cofactor/metal/ligand/drug binding for a protein wh…

Retrieve AlphaFold MUTAGEN annotations for a given UniProt accession. Returns experimental mutagenesis data mapped onto protein structures from UniProt. The qualifier must be a UniProt ACCESSION (e.g., 'P69905'). Note: Not all proteins have MUTAGEN annotations available in the database. As of 2026-05 AFDB's MUTAGEN annotation endpoint returns empty for all tested UniProts; the tool returns a structured error in that case.

Retrieve full AlphaFold 3D structure predictions for a given protein. IMPORTANT: The qualifier must be a UniProt ACCESSION (e.g., 'P69905' for HBA_HUMAN). Do NOT use UniProt entry names like 'HBA_HUMAN' - they will cause API errors. To find UniProt accession from a gene/protein name, use `UniProt_search` (e.g., query='gene:HBA' organism='human') or `UniProt_id_mapping` for ID conversion. Returns residue-level metadata including sequence, per-residue confidence scores (pLDDT), and structure down…

Retrieve summary details of AlphaFold 3D models for a given protein. IMPORTANT: The qualifier must be a UniProt ACCESSION (e.g., 'Q5SWX9' for MEIOB_HUMAN). Do NOT use UniProt entry names like 'MEIOB_HUMAN' - they will cause API errors. To find UniProt accession from a gene/protein name, use `UniProt_search` (e.g., query='gene:MEIOB' organism='human') or `UniProt_id_mapping` for ID conversion. Returns lightweight information such as sequence length, coverage, confidence scores, experimental meth…

Search scientific papers with BGPT and return structured, full-text-derived study evidence for critical appraisal. Unlike PubMed/EuropePMC/OpenAlex (which return titles and abstracts), each result includes methods and experimental techniques, sample size and population, results and conclusions, study context, paper limitations and biases, conflict of interest, data/code availability, software/tools used, quality scores, study blind spots, and a 'how_to_falsify' statement. Use it when an agent n…

Download a complete FBA-ready COBRA metabolic model from BiGG. Unlike BiGG_get_model (which returns only metadata counts), this returns the full constraint-based model ready for flux balance analysis (FBA): every reaction with its flux lower_bound/upper_bound and objective_coefficient, every metabolite with stoichiometry, all gene-protein-reaction associations, and compartments. The default JSON format returns the parsed COBRA model dict (loadable by COBRApy via cobra.io.load_json_model); forma…

Get BiGG Models database version and last update date. Returns BiGG version number, API version, and timestamp of last database update. Use for: checking data currency, version control, documentation, reproducibility. Check this to ensure using latest metabolic models data.

Get detailed information for a specific metabolic model. Returns organism, genome accession, reaction/metabolite/gene counts, and publication reference (PMID). Use for: understanding model scope, checking model currency, finding publications, model selection. Essential metadata before downloading or analyzing a model. Example model IDs: iJO1366 (E. coli), iMM904 (S. cerevisiae), Recon3D (human).

Get all reactions in a metabolic model. Returns reaction IDs, names, and organisms for all biochemical reactions in the model. Use for: analyzing metabolic capabilities, finding specific pathways, reaction enumeration, model comparison. Follow with BiGG_get_reaction for detailed reaction information including stoichiometry and gene-reaction rules.

Get detailed reaction information including stoichiometry, metabolites, gene-reaction rules, and database links. Query from specific model or universal database. Returns reaction equation, participating metabolites with coefficients, GPR (gene-protein-reaction) associations, bounds, and cross-references (KEGG, MetaCyc, BRENDA). Use for: understanding reaction mechanisms, gene associations, flux constraints, pathway analysis. Essential for metabolic modeling.

Get list of all genome-scale metabolic models in BiGG database (85+ models). Returns model IDs, organisms, reaction counts, metabolite counts, and gene counts for bacteria, archaea, and eukaryotes. Models cover E. coli, yeast, human, and other organisms. Use for: discovering available models, finding organism-specific models, comparing model sizes, metabolic modeling projects. Essential first step for flux balance analysis.

Search BiGG database for models, reactions, metabolites, or genes. Returns matching entries with IDs, names, and organism info. Search types: models (find organisms), reactions (find biochemical reactions), metabolites (find compounds), genes (find gene names/IDs). Use for: discovery, finding specific entities, cross-model searches, identifying pathways. Flexible search across entire database.

Get binding affinity data (Ki, IC50, Kd) for a single protein by UniProt ID. Returns SMILES structures and measured affinities from BindingDB. Essential for drug discovery target analysis. Example: P00533 (EGFR) returns known inhibitors with binding constants.

Get binding affinity data for multiple proteins by UniProt IDs. Returns SMILES and affinities for all matching ligands. Use for comparing binding profiles across protein family members or paralogs.

Find protein targets for a compound by SMILES structure. Returns proteins with binding affinity data for similar compounds. Essential for polypharmacology analysis and off-target prediction.

Phenome-wide association study (PheWAS) lookup for a single variant in BioBank Japan (BBJ), the largest Japanese-ancestry biobank. Returns every phenotype associated with the variant (the inverse of a GWAS), sorted by p-value. Accepts an rsID (auto-resolved to GRCh37 via Ensembl) or an explicit chr:pos:ref:alt variant in GRCh37. Use for cross-ancestry replication of a genetic association in an East Asian (Japanese) population.

Get detailed metadata for a specific Bioconductor R package by name. Returns comprehensive information including version, description, dependencies, biocViews categories, author/maintainer, license, download statistics, GitHub stars, reverse dependencies (packages that use it), and git repository URL. Useful for evaluating packages before installation or for understanding the Bioconductor ecosystem. Use Bioconductor_search_packages first to find package names.

Search Bioconductor bioinformatics R packages by keyword, topic, or analysis type. Bioconductor provides 2200+ open-source R packages for bioinformatics and computational biology covering RNA-seq, single-cell, proteomics, epigenomics, metabolomics, flow cytometry, imaging, and more. Returns matching packages with title, description, usage count, and relevance score. Powered by R-universe search index. Use this to find packages for specific bioinformatics workflows, then use Bioconductor_get_pac…

Get detailed information about a specific BioImage Archive study by accession number. Returns comprehensive metadata including title, description, imaging methods, organism, and file information. Example: 'S-BIAD634' returns an annotated fluorescence image dataset.

List the file/image manifest of a BioImage Archive study, giving per-file download paths and per-image experimental annotations. Unlike BioImageArchive_get_study (which returns study-level metadata), this returns the actual files: image Name, relative download path (append to the study's FTP/HTTPS root), file size, and any annotation columns the submitters attached to each image (e.g. staining, cells, treatment, channels, timepoint). Use limit/offset to page through large manifests; metadata.re…

Search the BioImage Archive specifically for biological image datasets (BioImages collection). This targets the BioImages-specific collection at EBI, returning imaging studies with microscopy-specific metadata. Example: 'cell division' finds imaging datasets of cell division processes.

Search the BioImage Archive (EBI BioStudies) for biological imaging datasets. The BioImage Archive hosts open microscopy, imaging, and visualization data from life sciences research. Search by imaging modality (cryo-EM, fluorescence, confocal), organism, technique, or scientific topic. Returns study accessions, titles, and metadata. Example: 'fluorescence microscopy cell' finds fluorescence imaging studies.

Search across 900+ biomedical ontologies in BioPortal (NCBO) for concepts matching a query. BioPortal is the world's largest repository of biomedical ontologies including GO (gene function), HPO (phenotypes), DOID (diseases), CHEBI (chemicals), UBERON (anatomy), SNOMED, MeSH, and 900+ more. Returns matching concepts with labels, IDs, definitions, and synonyms. Optionally filter by specific ontologies. Example: search 'melanoma' to find disease terms across DOID, NCIT, MESH, SNOMED; search 'apop…

Get cross-registry prefix mappings for a Bioregistry prefix. Maps a prefix (e.g. 'chebi', 'go', 'mondo') to its equivalent prefix/identifier in other registries: OBO Foundry, MIRIAM/identifiers.org, OLS, Name-to-Thing (N2T), BioPortal, FAIRsharing, BioContext, AberOWL, BARTOC, Wikidata, and more. Use this to translate a registry prefix between metadata systems (distinct from resolve_reference, which maps a specific data identifier to provider URLs). The standard Bioregistry_get_registry tool om…

Get metadata for a biological database/ontology by its Bioregistry prefix. Returns name, description, homepage, ID pattern, URI format, example IDs, and synonyms. Covers 2600+ resources including UniProt, ChEBI, GO, PDB, Ensembl, HGNC, and many more.

Resolve a compact identifier (prefix:id) to provider URLs across 2600+ biological databases. Given a database prefix (e.g., 'uniprot', 'chebi', 'go', 'pubmed') and an identifier, returns URLs from all known providers. Useful for cross-database linking and identifier resolution.

Search the Bioregistry for databases, ontologies, and resources by keyword. Use to find the correct prefix for identifier resolution, or to discover databases related to a topic (e.g., 'protein', 'gene ontology', 'metabolite').

Get full metadata for a specific bioRxiv or medRxiv preprint by its DOI. Returns comprehensive metadata including title, authors, abstract, publication date, version, category, corresponding author, funding information, and URLs (HTML, PDF, XML). Response time: <1 second. Use when you have a bioRxiv DOI (format: 10.1101/YYYY.MM.DD.XXXXXX). To SEARCH for preprints by keywords, use EuropePMC_search_articles with 'SRC:PPR' filter (fast, relevance-ranked), or web_search with 'site:biorxiv.org'. bio…

List recent bioRxiv or medRxiv preprints by date range. Returns up to 100 preprints posted between the given start and end dates, with title, authors, abstract, DOI, category, and publication status. The bioRxiv API returns results in batches of up to 100; use the cursor parameter for pagination (0 for first page, 100 for second, etc.). Use this to monitor new preprints in a date range or discover recent work in a field. For searching by DOI, use BioRxiv_get_preprint instead. For searching by k…

Get detailed information about a specific bioinformatics tool from the ELIXIR Bio.tools registry by its biotoolsID. Returns comprehensive metadata including name, description, homepage, tool type (command-line, web app, database, etc.), EDAM topics and operations, supported input/output data formats, programming languages, operating systems, license, documentation links, and credits. Use biotoolsID values discovered via BioTools_search. Example: 'samtools' for SAMtools, 'blast' for NCBI BLAST.

Search the ELIXIR Bio.tools registry for bioinformatics tools by EDAM operation (what the tool does). Returns tools that perform a specific computational operation. Useful operations include: 'Sequence alignment', 'Structure prediction', 'Variant calling', 'Gene expression analysis', 'Phylogenetic inference', 'Protein function prediction', 'Molecular docking', 'Genome annotation', 'Visualisation', 'Data retrieval', 'Statistical calculation', 'Clustering'. Use this to find software for a specifi…

Search nucleotide sequences using NCBI BLAST blastn against nucleotide databases. Requires Biopython (pip install biopython). IMPORTANT: This tool uses NCBI's remote BLAST service which is inherently SLOW - expect 5-30 minutes per query depending on server load. This is normal NCBI behavior, not a bug. For faster results, use smaller databases like 'refseq_select_rna' instead of 'nt'.

Search protein sequences using NCBI BLAST blastp against protein databases. Requires Biopython (pip install biopython). IMPORTANT: This tool uses NCBI's remote BLAST service which is inherently SLOW - expect 5-30 minutes per query depending on server load. This is normal NCBI behavior, not a bug. For faster results, use 'swissprot' or 'pdb' databases instead of 'nr'.

Find BMRB NMR entries associated with a PDB structure. Returns BMRB entry IDs linked to a given PDB accession, with match type (BLAST or author-provided). Useful for finding NMR chemical shift data for a protein with known crystal structure.

Get the citation/publication information for a BMRB NMR entry in BibTeX format. Returns author list, title, journal, DOI, and year. Useful for citing NMR data sources.

Get the BMRB-computed NMR assignment validation report for an entry (AVS and PANAV analyses). Unlike fetching raw deposited shifts, this returns per-residue assignment-consistency scoring: AVS (Assignment Validation Software) flags anomalous vs consistent backbone/sidechain assignments, and PANAV (Probabilistic Approach for protein NMR Assignment Validation) reports reference chemical-shift offsets and lists deviant/suspicious assignments. Result is keyed by the entry ID and contains 'avs' and…

Search BMRB (Biological Magnetic Resonance Data Bank) entries by keyword or molecule name. Returns matching BMRB entries with their IDs and summary information. Use the 'database' parameter to search either macromolecules (proteins, nucleic acids) or metabolomics small molecules.

Search for biological identifiers by name using BridgeDb. Enter a gene name, metabolite name, or protein name and get matching identifiers from supported databases. Useful for finding database-specific IDs when you only know the common name. Searches across all database sources available for the specified organism. Returns identifier-database pairs. For example, searching 'glucose' returns matches from MetaCyc, KEGG, and other databases.

Get detailed genome information from BV-BRC (Bacterial and Viral Bioinformatics Resource Center, formerly PATRIC). Returns genome metadata including organism name, taxonomy, genome size, GC content, assembly status, host, disease, and isolation information. BV-BRC is the primary NIAID-funded resource for bacterial and viral pathogen genomics. Example: '83332.12' returns Mycobacterium tuberculosis H37Rv.

Get detailed pathogen protein structure information from BV-BRC by PDB ID. Returns structure metadata including title, resolution, method, gene, organism, UniProt accession, PMID references, sequence, and taxonomic lineage. Useful for structural biology research on pathogen proteins. Example: '6VSB' returns the SARS-CoV-2 spike glycoprotein structure.

Get detailed taxonomy information for a pathogen from BV-BRC by NCBI Taxonomy ID. Returns taxon name, rank, lineage, alternative names, and genome counts. Useful for resolving pathogen taxonomy and understanding classification hierarchy. Example: '2697049' returns SARS-CoV-2 taxonomy with full lineage from Viruses to species.

Search for antimicrobial resistance (AMR) phenotype data in BV-BRC. Returns resistance/susceptibility data for pathogens against specific antibiotics, including MIC values, testing methods, and evidence. Essential for AMR surveillance research. Filter by 'antibiotic' (e.g. 'methicillin'), 'resistant_phenotype', and/or a specific BV-BRC 'genome_id'. There is no organism/species parameter: to scope results to one organism, first obtain its genome_id(s) (e.g. via BVBRC_search_genome_features / BVB…

Search for pathogen epitopes (B-cell and T-cell) in BV-BRC. Returns epitope sequences, types, protein associations, and assay results for vaccine and immunology research. Covers epitopes from IEDB for bacteria and viruses including SARS-CoV-2, influenza, tuberculosis, and other pathogens. Example: search taxon_id 2697049 for SARS-CoV-2 spike glycoprotein epitopes.

Search for genome features (genes, proteins, CDS) in BV-BRC pathogen genomes. Find specific genes like antimicrobial resistance genes (mecA, blaTEM), virulence factors, or any annotated feature across thousands of pathogen genomes. Example: search for 'mecA' gene (methicillin resistance) across all genomes.

Search for pathogen genomes in BV-BRC by organism name or keyword. Returns a list of matching genomes with metadata. Useful for finding bacterial or viral genome assemblies. Example: 'Staphylococcus aureus' returns all S. aureus genome assemblies.

Search for metabolic pathways in pathogen genomes from BV-BRC. Returns KEGG pathway assignments with EC numbers, enzyme descriptions, and genome associations. Useful for understanding pathogen metabolism and identifying potential drug targets. Example: search for pathways in M. tuberculosis (taxon_id 1773).

Search for pathogen protein structures in BV-BRC by organism (taxon_id), gene name, or experimental method. Returns PDB entries with titles, resolution, methods, and references. Covers viral and bacterial protein structures linked to BV-BRC genome annotations. Example: search for all SARS-CoV-2 protein structures with taxon_id 2697049.

Search for specialty genes in BV-BRC including virulence factors, antibiotic resistance genes, drug targets, and human homologs across pathogen genomes. Data sourced from CARD, NDARO, VFDB, PATRIC_VF, and other curated databases. Essential for identifying key pathogenicity and resistance determinants. Example: search for 'mecA' (methicillin resistance gene in Staphylococcus).

Search for functional subsystems (curated groups of functionally related proteins) in BV-BRC pathogen genomes. Returns subsystem classifications including superclass, class, subclass, and specific functional roles. Useful for understanding pathogen functional organization and comparing metabolic capabilities. Example: search for virulence-related subsystems in M. tuberculosis.

Search influenza and pathogen surveillance data in BV-BRC. Returns surveillance samples with host species, geographic location, subtype, and collection metadata. Essential for epidemiological tracking of influenza subtypes (H5N1, H1N1, H3N2) and other monitored pathogens across global regions. Example: search for H5N1 cases in North America.

Search for pathogen taxonomy entries in BV-BRC by keyword. Returns matching taxa with names, ranks, genome counts, and lineage information. Useful for discovering pathogen taxon IDs for use with other BV-BRC tools. Example: 'Mycobacterium' returns the Mycobacterium genus and all its species.

Get CATH protein structure superfamily information by CATH ID. CATH classifies protein domain structures into a four-level hierarchy: Class, Architecture, Topology, Homologous superfamily. Superfamilies group evolutionarily related domains. Returns classification name, number of families, and domain counts. Example: 2.40.50.140 returns 'Nucleic acid-binding proteins' with 227 S35 families and 2,879 total domains. Note: 'classification_name'/'classification_description' are null for some superfa…

List functional families (FunFams) within a CATH superfamily. FunFams group protein domains that share the same function, identified by conserved specificity-determining residues. Returns FunFam numbers, names, member counts, and representative domain IDs. Example: superfamily 1.10.510.10 (Globin-like) has 4,524 FunFams including CaM kinase (2,698 members) and Myosin light chain kinase (1,234 members). Use CATH_get_superfamily to find superfamily IDs first.

Get list of cancer studies from cBioPortal

Get detailed information about a specific gene by Entrez Gene ID. Returns gene symbol, aliases, type, and chromosome location.

Get all genes in a specific gene panel. Essential for understanding what genes are covered when analyzing mutation frequency data from targeted sequencing studies.

Get all gene panels used in cBioPortal studies. Gene panels define which genes were sequenced in a study. Use to understand coverage before querying mutation data.

Search for genes by keyword or query (gene symbol or alias). Alias: cBioPortal_search.

Get molecular profiles for a cancer study. Molecular profiles include mutation data, copy number alterations, mRNA expression, etc.

Get mutation data for specific genes in a cancer study. This uses the updated cBioPortal API that requires molecular profiles.

Get all patients in a cancer study

Get all sample lists (cohort definitions) in a cancer study. Sample lists define groups like 'samples with mutations', 'samples with CNA data'. Use for targeted queries.

Get all samples in a cancer study

Search for datasets on Data.CDC.gov (CDC's Socrata-based open data portal). Returns a list of available datasets matching search criteria. Use this to discover datasets before querying data.

Convert natural language queries to Solr syntax for Cellosaurus API searches. Uses semantic similarity to map terms to appropriate fields.

Get detailed information about a chemical entity from ChEBI (Chemical Entities of Biological Interest) by its CHEBI ID. Returns the compound name, definition, synonyms, molecular formula, mass, InChIKey, SMILES, charge, and star rating (curation quality: 3=manually curated, 2=checked, 1=unchecked). ChEBI is maintained by EBI and is the standard reference for small molecules in biology. Example: CHEBI:15365 returns acetylsalicylic acid (aspirin).

Get the ontology children (is-a, has-part, has-functional-parent, etc.) of a ChEBI compound. ChEBI uses a rich ontology with relationships like 'is a', 'has part', 'has role', 'has functional parent', and 'is conjugate base/acid of'. This enables navigation of the chemical classification hierarchy. Example: CHEBI:15365 (aspirin) children include NCX-4040 (has functional parent), acetylsalicylate (conjugate base), and Yosprala (has part).

Get the ontology parents (is-a / has-role ancestors) of a ChEBI compound, navigating UPWARD in the chemical classification hierarchy (the symmetric counterpart of ChEBI_get_ontology_children). ChEBI uses a rich ontology with relationships like 'is a', 'has role', 'has functional parent', and 'is conjugate base/acid of'. Example: CHEBI:15377 (water) parents include 'mononuclear parent hydride' and 'inorganic hydroxy compound' (is a), and 'human metabolite', 'amphiprotic solvent' (has role).

Search the ChEBI database for chemical entities by name, formula, or keyword. Uses Elasticsearch-powered full-text search across compound names, synonyms, definitions, and formulas. Returns matching compounds with IDs, names, formulas, and masses. Note: smiles and inchikey are not available from the search endpoint (use ChEBI_get_compound for structural data). Compound names are HTML-stripped. ChEBI contains 195,000+ chemical entities of biological interest.

Get detailed information about a target (protein, gene, etc.) by its ChEMBL target ID. To find a target ID, use ChEMBL_search_targets with a target name or gene symbol.

Get all assays associated with a target by ChEMBL target ID. To find a target ID, use ChEMBL_search_targets with a target name or gene symbol.

Search activity data by molecule, target, assay, or activity values. Supports filtering by IC50, Ki, EC50, etc. Note: activity data is sparse and depends on curation; some specific filter combinations can legitimately return zero results. For reliable retrieval, prefer starting from a known target (`target_chembl_id__exact`) or assay (`assay_chembl_id__exact`). To find molecule or target IDs, use ChEMBL_search_molecules or ChEMBL_search_targets.

Search assays by various criteria including assay type, target, organism. Use this tool to find assay ChEMBL IDs needed for other tools.

Search ATC (Anatomical Therapeutic Chemical) classifications for drugs. To find a drug ID, use ChEMBL_search_drugs.

Search binding sites by target, site name, or other criteria.

Search cell lines used in ChEMBL assays.

Search compound structural alerts (toxicity warnings) by molecule or alert type.

Search ChEMBL documents (publications) by various criteria.

Search drugs by name, approval status, or other criteria. Use the `query` parameter to search by drug name (e.g., {"query": "sotorasib"}). Use `max_phase` to filter by development phase.

Search molecules by name, ChEMBL ID, or other criteria. Supports filtering and pagination. Note: some text-based filters (especially `pref_name__contains`) may legitimately return zero results because `pref_name` coverage is incomplete and may not contain common drug names. For reliable retrieval, prefer ID-based filters like `molecule_chembl_id` or use `ChEMBL_get_molecule` when you already have a ChEMBL ID.

Search for molecules similar to a given SMILES, chembl_id, or compound or drug name, using the ChEMBL Web Services. Note: This tool is designed for small molecule compounds only. Biologics (antibodies, proteins, oligonucleotides, etc.) do not have SMILES structures and cannot be used for similarity search. For biologics similarity search, use BLAST_protein_search (requires amino acid sequence) or UniProt_search. For small molecules, use PubChem_search_compounds_by_similarity (requires SMILES in…

Search molecules similar to a given SMILES string using Tanimoto similarity. Returns molecules with similarity above the threshold.

Search molecules by substructure using SMILES. Returns molecules containing the specified substructure.

Search targets by name, organism, target type, or other criteria. Use this tool to find target ChEMBL IDs needed for other tools.

Search tissues used in ChEMBL assays.

Perform enrichment analysis to identify transcription factors and histone modifications enriched in your data. **Required: Provide ONE input type** - (1) BED genomic regions, (2) DNA sequence motif (IUPAC notation: A/T/G/C/W/S/M/K/R/Y/B/D/H/V/N), or (3) gene symbol list. Compares your input against 433,000+ ChIP-seq/ATAC-seq/Bisulfite-seq experiments to identify significant enrichment. Returns ranked list of proteins bound to your regions/motif or regulating your genes. Note: Returns submission…

ChIP-Atlas Colocalization: for a given antigen/transcription factor in a specific tissue class, return the other proteins whose ChIP-seq peaks co-occur (co-bind) genome-wide, ranked by overlap score. Reveals protein complexes and co-regulators (e.g. CTCF in Blood co-binds the cohesin subunits RAD21, SMC1A, SMC3, STAG1 and ZNF143). Requires both an antigen and a cell-type/tissue class because the matrix is precomputed per antigen-per-tissue. To discover which tissue classes exist for an antigen,…

Search ChIP-Atlas experiment metadata including experiment IDs, antigens, cell types, and processing statistics. Filter by genome, antigen (TF/histone), or cell type. Returns experiment details with SRX/ERX/DRX IDs. Use for: finding relevant experiments, getting experiment metadata, dataset discovery, batch analysis planning.

Get download URLs for ChIP-Atlas peak-call data in BigWig, BED, or BigBed format. BigWig contains coverage scores in RPM (Reads Per Million). BED/BigBed contain peak regions with MACS2 (Model-based Analysis of ChIP-Seq) peak caller scores. Use for: downloading raw data, integrating with genome browsers (UCSC, IGV), custom analysis, visualization.

Get detailed information about a specific assertion in CIViC database by assertion ID. Assertions are higher-level clinical interpretations that integrate multiple evidence items into formal statements about clinical actionability. Returns assertion description, evidence summary, and associated molecular profile or variant information.

Get detailed information about a specific evidence item in CIViC database by evidence ID. Evidence items link molecular profiles or variants to clinical outcomes, therapies, or disease states with curated literature support. Returns evidence description, level (A-E), type (PREDICTIVE, DIAGNOSTIC, PROGNOSTIC), significance, direction, status, disease, therapies, molecular profile, and citation.

Get detailed information about a specific molecular profile in CIViC database by molecular profile ID. Molecular profiles represent combinations of variants or features (e.g., 'BRAF V600E') that serve as biomarkers for clinical interpretation. Returns profile name, description, clinical significance score, and associated evidence items with diseases, therapies, and evidence levels.

Get detailed information about a specific variant in CIViC database by variant ID. Variants represent specific genetic alterations (SNVs, indels, structural variants, etc.) with clinical significance in cancer. Returns variant name, coordinates, gene association, and related information.

Get all variants associated with a specific gene in CIViC database. Returns variant information including names, coordinates, and associated evidence. Use gene_name or gene_symbol (e.g., 'EGFR', 'BRAF') for convenience, or gene_id if already known. Uses cursor-based pagination to return up to 500 variants per call.

Search for assertions in CIViC database. Assertions are higher-level clinical interpretations that integrate multiple evidence items into formal statements about clinical actionability. Returns a list of assertions with descriptions and associated molecular profiles or variants. Filter by therapy (e.g., 'ponatinib'), disease (e.g., 'leukemia'), or variant name.

Search for diseases in CIViC database. Returns a list of cancer diseases and conditions (with IDs and names) that are associated with variants and evidence in CIViC. Use 'query' or 'name' to filter by disease name (e.g., 'leukemia', 'melanoma').

Search for evidence items in CIViC database. Evidence items are curated statements linking variants or molecular profiles to clinical outcomes, therapies, or disease states. Returns evidence with disease, therapies, significance, direction, and citation. Filter by therapy name (e.g., 'imatinib'), disease name (e.g., 'leukemia'), or molecular profile name (e.g., 'BRAF V600E').

Search for genes in CIViC (Clinical Interpretation of Variants in Cancer) database. CIViC is a community knowledgebase for expert-curated interpretations of variants in cancer. Returns genes with their IDs, names, and descriptions. Use this to find genes of interest before querying variants or evidence.

Search for molecular profiles in CIViC database. Molecular profiles represent combinations of variants or features (e.g., 'BRAF V600E', 'EGFR T790M') that serve as biomarkers for clinical interpretation. Returns a list of molecular profiles with IDs and names. Use the 'query' parameter to filter by profile name (alias: 'name').

Search for therapies (drugs/treatments) in CIViC database. Returns a list of cancer therapies and drugs (with IDs and names) that are associated with variants and evidence in CIViC. Use 'query' or 'name' to filter by therapy name (e.g., 'imatinib', 'pembrolizumab').

Search for variants in CIViC database. Returns a list of variants with their IDs and names. Use 'query' to search by variant name (e.g., 'V600E'), or 'gene'/'gene_name' to get all variants for a specific gene (e.g., 'EGFR'). To find a specific variant in a gene, combine gene_name + variant_name (e.g., gene_name='EGFR', variant_name='L858R'). Note: CIViC stores variants without gene prefix — do not combine gene and variant in a single string.

Classify a chemical structure into the ClassyFire ChemOnt chemical taxonomy by its InChIKey. Returns the hierarchical classification (kingdom -> superclass -> class -> subclass -> direct parent), intermediate ontology nodes, molecular framework, structural substituents, and a textual description. Use to assign a compound to a standardized chemical class for grouping, filtering, or annotation. Requires a full InChIKey (compute it from a structure via OPSIN/PubChem/RDKit first). Only previously-c…

Get ClinGen dosage sensitivity curation for a specific gene using the JSON API. Returns haploinsufficiency (HI) and triplosensitivity (TS) assertion scores with genomic coordinates in both GRCh37 and GRCh38. Scores: 0=No evidence, 1=Little evidence, 2=Emerging evidence, 3=Sufficient evidence, 30=Gene associated with autosomal recessive phenotype, 40=Dosage sensitivity unlikely. Also includes pLI score, HI prediction score, and associated diseases with MONDO IDs.

Search ClinGen dosage sensitivity curations by genomic region. Returns all genes and recurrent CNV regions with dosage sensitivity curations overlapping the specified chromosomal coordinates. Supports both GRCh37 and GRCh38. Results include haploinsufficiency and triplosensitivity scores. Useful for interpreting the clinical significance of deletions and duplications detected by chromosomal microarray.

Look up a genetic variant in the ClinGen Allele Registry by HGVS notation. Returns the canonical allele identifier (CA ID), community standard title, and cross-references to ClinVar, dbSNP, COSMIC, gnomAD, and other databases. The Allele Registry normalizes variant representations across different coordinate systems and nomenclatures. Example: 'NM_000546.6:c.743G>A' returns CA000387 (TP53 p.Arg248Gln) with links to ClinVar, COSMIC, and gnomAD.

Reverse-lookup the ClinGen Allele Registry by an external identifier. Given a dbSNP rsID (dbsnp_rs) or a ClinVar VariationID (clinvar_variation_id), resolves it to the canonical ClinGen allele (CA id) and returns ALL cross-references (ClinVar alleles/variations, dbSNP, COSMIC, gnomAD, MyVariantInfo) plus the community standard HGVS title and genomic allele HGVS. This complements ClinGenAR_lookup_allele (which takes an HGVS expression): use this when you start from an rsID or ClinVar variation i…

CHA2DS2-VASc score: stroke risk in non-valvular atrial fibrillation, used to decide oral anticoagulation. Deterministic point score (0-9). Provide age plus boolean risk factors. Returns score, anticoagulation interpretation, and per-component points.

Child-Pugh score and class (A/B/C): severity of chronic liver disease / cirrhosis and surgical risk. Sum (5-15) of bilirubin, albumin, INR, ascites, and encephalopathy points. Provide lab values and clinical grades.

CURB-65 score: community-acquired pneumonia severity and disposition (outpatient vs inpatient vs ICU). Point score (0-5) from confusion, urea, respiratory rate, blood pressure, and age >=65.

Estimated glomerular filtration rate (eGFR) by the CKD-EPI 2021 creatinine equation (race-free, Inker 2021), with CKD stage (G1-G5). Inputs: serum creatinine, age, sex. Reported in mL/min/1.73m^2.

HAS-BLED score: 1-year major bleeding risk in patients with atrial fibrillation on anticoagulation. Point score (0-9); >=3 indicates high risk warranting caution and review of reversible factors. Complements CHA2DS2-VASc.

MELD-Na score (UNOS/OPTN 2016): 90-day mortality risk in chronic liver disease and liver transplant allocation priority. Range 6-40. Uses creatinine, bilirubin, INR, and sodium (with the standard lower/upper bounds and dialysis rule).

qSOFA (quick SOFA): bedside identification of patients with suspected infection at higher risk of poor outcome. Point score (0-3); >=2 is concerning for sepsis. Uses respiratory rate, mentation, and systolic blood pressure only.

Wells score for deep vein thrombosis (DVT) pretest probability. Sum of clinical features (minus 2 if an alternative diagnosis is as likely); score >=2 means DVT likely. Guides D-dimer vs ultrasound workup.

Wells score for pulmonary embolism (PE) pretest probability. Weighted score giving three-tier (low/moderate/high) and two-tier (PE unlikely <=4 / likely >4) risk to guide D-dimer vs CT pulmonary angiography.

Get aggregate statistics about the ClinicalTrials.gov database, including total number of registered studies and size distribution. Returns database-wide counts and file size metrics.

Get full details for a specific clinical trial by NCT ID from ClinicalTrials.gov. Returns comprehensive protocol including title, objectives, eligibility criteria, interventions, outcome measures, contacts, and results summary. Example: NCT04280705 returns the ACTT remdesivir COVID-19 trial with all details.

Search ClinicalTrials.gov for all clinical trials testing a specific drug, biologic, device, or other intervention. Returns trial IDs, titles, phases, status, and conditions being studied. Ideal for finding all trials for a given drug (e.g., 'nivolumab') across all disease indications.