38 providers
Life sciences data for AI agents
Look up proteins, genes and expression data, 3D structures and bioactive compounds. Search clinical trials, regulatory documents and the biomedical literature.
Life sciences providers 38
Every one callable through the same Firecrawl API key.
Research IndexFind scientific papers, inspect metadata, read relevant passages and follow citation relationships.AlphaFold Protein Structure DatabaseAlphaFold DB, EMBL-EBI and Google DeepMind's database of over 260 million predicted protein structures, through its keyless API at alphafold.ebi.ac.uk: the predicted structure of a protein with its pLDDT confidence, the residue ranges that are confident and those likely disordered, and links to the model, PAE and confidence files; predicted complexes containing a protein with their partners and interface scores (ipTM, ipSAE, pDockQ); and AlphaMissense pathogenicity predictions for every possible missense substitution in a human protein. Takes UniProt accessions, entry names, AlphaFold model IDs, entry URLs or gene symbols. Data is CC BY 4.0: every result carries its licence, citations and entry page.
France public medicines database (BDPM)France's base de données publique des médicaments (BDPM), run by the ANSM with the HAS and the national health insurance fund: every medicine marketed in France now or within the last two years (about 15,900), with its authorisation, composition and packs, each pack's reimbursement rate, price and public price, generic groups, prescribing conditions and shortage notices, and every HAS Transparency Committee opinion since 2002 with its SMR (clinical benefit) and ASMR (improvement in clinical benefit) levels and wording. Search medicines, read one medicine whole, or list HAS opinions by date and level. Values are the BDPM's French text, verbatim; statuses and SMR/ASMR levels carry English glosses. Data is under the Etalab Open Licence; every answer carries the source mention and the update date the licence requires.
ClinicalTrials.govClinicalTrials.gov study registry via the public v2 REST API: search studies by condition, intervention, sponsor, location, status or Essie expression with page-token pagination; fetch one study record by NCT number; batch-resolve the raw ACTUAL/ESTIMATED start, primary-completion and completion dates for known NCT numbers; autocomplete conditions, interventions, sponsors and facilities; and read registry-wide field value statistics.
ClinVarClinVar, NCBI's public archive of human genetic variants and their clinical classifications (4.7 million records), through NCBI E-utilities. Search variants by gene, HGVS name, rsID, canonical SPDI, condition or free text, filtered by germline classification, review status or star rating, variant type and molecular consequence; read one variant by VCV accession, variation ID, ClinVar URL, rsID or HGVS with its aggregate germline classification, somatic clinical impact and oncogenicity, review status and stars, conditions with MedGen, OMIM, MONDO, Orphanet and HPO IDs, per-condition RCV aggregates, and each submitter's classification with criteria, allele origin, method and cited PMIDs. Every output carries ClinVar's citation and its clinical-use disclaimer.
CMS Medicare Drug Prices: Negotiated Prices and Part B ASP Payment LimitsMedicare drug prices from two CMS sources. First, the drugs CMS selected for Medicare drug price negotiation under the Inflation Reduction Act and their negotiated prices, which the statute calls maximum fair prices (MFPs), from CMS's Selected Drug List and Negotiated Prices file: the 10 drugs with prices from 2026 (Eliquis, Jardiance, Xarelto, Januvia, Farxiga, Entresto, Enbrel, Imbruvica, Stelara, NovoLog/Fiasp), the 15 from 2027 (among them Ozempic, Rybelsus and Wegovy) and the 15 selected for 2028, each with its MFP per 30-day equivalent supply by period, status, Part B HCPCS codes, deselection and every NDC-11 with its per-unit MFP. Second, Medicare Part B payment allowance limits for physician-administered drugs (generally 106% of the manufacturer-reported average sales price, ASP) per HCPCS Level II code from CMS's quarterly ASP pricing files since 2021, for example Keytruda (J9271) or Opdivo (J9299), joined to the ASP NDC-HCPCS crosswalk for each code's NDCs, labelers and billable units per package, plus one code's limit across every quarter. CPT-coded rows (AMA copyright) are left out. Public domain U.S. Government data; every result carries CMS's attribution.
CDC PLACES local health estimates (data.cdc.gov)CDC PLACES (Population Level Analysis and Community Estimates) on data.cdc.gov: model-based estimates of the share of adults with 40 measures, from chronic diseases (diabetes, obesity, high blood pressure, high cholesterol, COPD, asthma, coronary heart disease, stroke, depression, arthritis, cancer) to risk behaviors (smoking, binge drinking, physical inactivity, short sleep), prevention (health insurance, checkups, screenings), health status, disabilities and social needs, for every county, place, census tract and ZIP Code Tabulation Area in the 50 states and DC. Releases from 2020 on; the latest uses BRFSS survey data from 2022 and 2023. Each estimate carries 95% confidence limits and total and adult population, with a derived count of adults and a summary over every matching location, the denominators patient-based epidemiology and drug forecasting start from. Public domain U.S. Government data; every result carries CDC's attribution and terms.
California HCAI prescription drug price reports (WAC increases and new drugs)California's prescription drug cost transparency reports (CTRx, SB 17) from the Department of Health Care Access and Information (HCAI), published on the CalHHS Open Data Portal. Manufacturers report wholesale acquisition cost (WAC, list price) increases of more than 16% over three years for drugs costing over $40 per course of therapy, with the increase, the new WAC, patent expiry, sales, their stated reasons and any acquisition, plus a five-year WAC history per report; and new drugs introduced at a WAC above the Medicare Part D specialty-drug threshold, with launch WAC, marketing and pricing plan, estimated patients and FDA breakthrough and priority review flags. The current year's monthly update and annual files back to 2019, searchable by drug, NDC, manufacturer and date, with derived increase percentages including the statute's three-year measure. CC BY; every result carries HCAI's attribution.
CMS data.cms.govCMS data.cms.gov: Provider Data Catalog datasets (NPI-keyed clinician procedure volume; Care Compare search and detail for hospitals, nursing homes, home health agencies and dialysis facilities); what Medicare Part D, Medicare Part B and Medicaid spend on each drug, searched by brand or generic name or ranked by spending, claims, patients or unit-cost growth; and the Medicare and Medicaid Spending by Drug collectors (Part D, Part B and Medicaid annual and quarterly spending by drug and manufacturer, Part D prescribing by geography and drug): dataset catalogue, version status, full paged listing pinned to one release, and single-record lookup. CMS data is a U.S. Government work; spending answers carry CMS's attribution line.
Medicaid open data: SDUD and NADACCMS Medicaid open data (data.medicaid.gov, DKAN): State Drug Utilization Data (Medicaid units, prescriptions and dollars reimbursed by state, NDC and quarter) and NADAC (weekly National Average Drug Acquisition Cost per unit by NDC), collected per dataset-year through the public datastore API under the shared collector contract (source status, pinned paginated listing, single-record lookup). Two agent-facing NADAC functions answer price questions directly: `prices/search` reads the newest weekly file by drug name or NDC (or the rates that took effect in a date range) and `prices/history` returns every rate one NDC has had since 2013, each with CMS attribution.
ChEMBLChEMBL, EMBL-EBI's open database of bioactive drug-like molecules (www.ebi.ac.uk/chembl), through its keyless JSON web services: molecule search by name or synonym and molecule records (structure, properties, max phase, synonyms, mechanisms of action), drug indications with EFO/MeSH terms and phases, drug targets with UniProt accessions, and bioactivity measurements (IC50, Ki, Kd, EC50, pChEMBL) by molecule and/or target. Data CC BY-SA 3.0 (attribution, share-alike); every output carries the licence and ChEMBL release.
European Medicines AgencyEU centralised medicines data from EMA's published data files: authorised medicines (EPAR pages), CHMP and CVMP opinions and withdrawn applications, applications under evaluation, EU orphan designations, EPAR and product documents, and EMA-assessed supply shortages, with EMA attribution on every result.
EU Clinical Trials (CTIS)The EU Clinical Trials Information System (CTIS), the European Medicines Agency's public register of every clinical trial authorised under the EU Clinical Trials Regulation (about 12,700 trials, including those transitioned from EudraCT), through its public JSON API: search by text, condition, medicine, sponsor, EU/EEA country, phase, status, decision or start date, rare disease, orphan designation and results, and read one full trial record by EU CT number, CTIS URL or a linked NCT, EudraCT, ISRCTN or WHO UTN number, with per-country decisions, statuses and sites, products and active substances, endpoints, eligibility, safety events, corrective measures, modifications and results. Every result carries EMA's reuse terms.
Europe PMCEurope PMC (europepmc.org), EMBL-EBI's open life-science literature database, through its public REST and Annotations APIs: search PubMed abstracts, open-access full text and preprints with Europe PMC query syntax, section-level and accession filters; read article records with abstracts, licences, grants, retraction status and preprint-to-publication links; list text-mined genes, diseases, chemicals, organisms and accessions grounded to database IDs; and read the full text of open-access articles and preprints whose licence allows commercial reuse (CC0, CC BY, CC BY-SA, CC BY-ND). Every result carries attribution to Europe PMC and each article's own licence.
FDAUS FDA drug, food and device enforcement recalls, Drugs@FDA applications and approval documents (letters, labels, reviews), complete response letters, orphan drug designations and approvals with exclusivity dates, Orange Book products with therapeutic equivalence codes, patents and exclusivities, the Purple Book of licensed biologics with reference product exclusivity and biosimilar and interchangeable status, the Paragraph IV patent certifications list, novel drug and first generic approvals, approved cell and gene therapies, structured labeling, device 510(k) decisions, PMA approvals and supplements, device classifications, GUDID device identifiers, MAUDE device and FAERS drug adverse-event reports, the UNII substance registry, the NDC directory, drug shortages, and warning letters, via openFDA, fda.gov, accessdata.fda.gov and the Purple Book.
G-BA early benefit assessments (AMNOG)Germany's Federal Joint Committee (G-BA) decisions on the added benefit of new medicines under § 35a SGB V (AMNOG), the assessment that sets the basis for price negotiations in Germany. Every decision in force (about 1,050 since 2011), from G-BA's machine-readable decision file: per patient group the extent of added benefit (erheblich, beträchtlich, gering, nicht quantifizierbar, nicht belegt, geringerer Nutzen) and its certainty with English glosses, the appropriate comparator therapy, ICD-10 and ATC codes, endpoint results, the rationale summary and requirements for quality-assured use. Search by drug, indication, ICD-10, ATC, comparator, benefit extent and decision date, or read one decision in full. Values stay in German. Patient numbers and annual therapy costs are not in G-BA's machine-readable file; the decision text on the linked g-ba.de page has them.
HGNC (HUGO Gene Nomenclature Committee)HGNC, the authority for human gene names, through the genenames.org REST API: resolve any gene symbol, alias, previous symbol or ID (HGNC, NCBI Gene, Ensembl, UniProt, RefSeq, MGI, RGD) to the current approved symbol, with ambiguous aliases listed rather than guessed; check up to 50 genes at once; and search genes by name, symbol pattern, cytoband or locus type. Records carry the approved name, aliases, previous symbols, locus type, location, gene groups and cross-references that join NCBI Gene, Ensembl, UniProt, OMIM and PubMed. HGNC data is CC0.
Population allele frequencies and gene constraint (Broad Institute)gnomAD, the Genome Aggregation Database of the Broad Institute, through the public GraphQL API behind its browser: how often a variant occurs in 807,162 people (v4.1) by sequencing type (exomes, genomes, both combined) and genetic ancestry group, with homozygote and hemizygote counts, quality filters, the grpmax filtering allele frequency and its consequence on the MANE Select transcript with LOFTEE's loss-of-function call; a gene's constraint (pLI, LOEUF with its confidence interval and percentile, observed and expected pLoF, missense and synonymous variants, z scores); and the variants gnomAD lists in a gene, filtered by consequence, LOFTEE call and allele frequency. Takes gene symbols, Ensembl IDs, gnomAD variant IDs, rsIDs, ClinGen allele IDs and browser URLs, on gnomAD v4.1 and v3.1.2 (GRCh38) or v2.1.1 (GRCh37). Primary data is CC0: every answer carries gnomAD's attribution, citation and terms of use.
GTEx Portal: gene expression by tissue and eQTLsThe GTEx Portal (Genotype-Tissue Expression project: Broad Institute, NIH Common Fund) through its public API v2, open-access aggregate data only. Median gene expression (TPM) across 54 human tissue sites from about 950 post-mortem donors, and significant single-tissue cis-eQTLs (variant, gene, tissue, normalized effect size, p-value) looked up by gene or by variant (rsID, GTEx variant ID or GRCh38 position). Releases GTEx Analysis v11 (default, GENCODE v47), v10 and v8. Outputs carry Ensembl, GENCODE, NCBI Gene and HGNC gene IDs, dbSNP rsIDs, GRCh38 and GRCh37 variant IDs and UBERON tissue IDs to join with UniProt, Open Targets, HPA, HGNC and NCBI Gene, and the GTEx Portal Data License attribution. No sample-level or donor-level data.
Health Canada Drug Product DatabaseHealth Canada's Drug Product Database (DPD), the register of drugs authorized for sale in Canada: about 58,000 human, veterinary, disinfectant and radiopharmaceutical products, approved, marketed, dormant or cancelled. Search by brand name, active ingredient, DIN, ATC code, AIG number or company, and read one product by DIN or drug code: active ingredients with strengths, dosage forms, routes, schedules, WHO ATC class, current status with its date and the original market date, the DIN owner with address, and the product monograph link. Data is under the Open Government Licence - Canada 2.0; every result carries its attribution statement.
ISRCTNISRCTN, the UK's primary clinical study registry and a WHO primary registry (run by BioMed Central, part of Springer Nature), through its public XML API: search about 29,000 interventional and observational studies by condition, intervention, sponsor, funder, title, recruitment country, phase, registration date or computed recruitment status, and read one full record by ISRCTN, URL or a linked NCT, EudraCT, CTIS, IRAS or CPMS number, with eligibility, outcomes, centres, sponsors, funders, secondary identifiers and linked publications (PMIDs, DOIs). Records are CC BY 4.0: every result carries its licence, the study DOI and a citation.
NHSBSA Secondary Care Medicines Data (SCMD)Medicines issued by NHS hospital trusts in England (England only: acute, teaching, specialist, mental health and community trusts), from the NHS Business Services Authority's Secondary Care Medicines Data (SCMD) on the NHSBSA Open Data Portal: monthly quantities of every dm+d Virtual Medicinal Product (VMP) issued by each trust from April 2019, with an indicative cost (list or reimbursement price, not the confidential price trusts paid). Finds a medicine by part of its product name or its VMP SNOMED code over up to 24 months, finalised months where NHSBSA has published them and provisional ones after, and totals it by product, by trust (with the trust's name from NHS England's ODS directory) or by month. Hospital issues only, not community prescribing. Open Government Licence v3.0; every answer carries NHSBSA's attribution and the cost caveat.
CMS Open Payments research paymentsResearch payments that drug and device manufacturers and group purchasing organizations reported to CMS under the Physician Payments Sunshine Act (Open Payments): about a million payments a year to teaching hospitals, research institutions and practitioners for clinical trials and other research, each with its sponsor, site, principal investigators, study name and ClinicalTrials.gov ID, products and amount. Search payments, rank them by sponsor, site, investigator, study or product, and read CMS's yearly totals nationally, per manufacturer or per teaching hospital. Program years 2019 to 2025 (published June 2026). Public domain U.S. Government data; every answer carries CMS's attribution, its disclaimer that a reported relationship is not necessarily improper, and its rule for disputed records.
Orphanet (Orphadata): rare disease prevalence, codes and genesOrphanet, the INSERM rare disease knowledge base, through its Orphadata Science API (CC BY 4.0). Finds rare diseases by name, synonym or abbreviation ("Pompe disease", "ALS"), ORPHAcode, orpha.net page, OMIM number or ICD-10/ICD-11 code, with definition, synonyms and Orphanet's mappings to ICD-10, ICD-11, OMIM, MONDO, MeSH, UMLS, MedDRA and GARD. Gives each disease's prevalence, birth prevalence, lifetime prevalence, annual incidence and case counts by area (worldwide, Europe and other regions, 130 countries) with Orphanet's prevalence class, mean value per 100,000, validation status and source PMIDs, plus inheritance and age of onset; and its gene associations with association type, HGNC, Ensembl, UniProt and OMIM identifiers. About 11,600 entities; 6,700 have prevalence data, 4,200 have genes. Orphadata releases twice a year; every answer carries its data version, Orphadata's citation and its AI-use notice.
Open Targets PlatformDrug-target evidence from the Open Targets Platform (EMBL-EBI, Wellcome Sanger Institute and partners; data CC0): target profiles with tractability, safety liabilities and cross-references; diseases and phenotypes across EFO, MONDO, HP and Orphanet with synonyms and vocabulary mappings; target-disease association scores broken down by evidence type and datasource; the evidence records behind them (GWAS, rare variants, somatic drivers, clinical precedence, literature); and known drugs and clinical candidates with mechanisms, clinical stage and trial or regulatory references. Targets resolve from symbols, synonyms and UniProt accessions, diseases from names and IDs, including retired EFO IDs. Uses the keyless Open Targets GraphQL API.
Human Protein AtlasThe Human Protein Atlas (KTH, Uppsala University and SciLifeLab, Sweden), CC BY 4.0: for every human protein-coding gene, where the protein sits in the cell from antibody immunofluorescence, with a reliability grade (Enhanced, Supported, Approved, Uncertain) and GO ID per location, the evidence behind Enhanced grades and single-cell variation; RNA tissue and single cell type specificity with the tissues and cells that express it most; protein evidence, protein classes and secretome annotation; and the antibodies HPA used, with providers, product names, clonality and per-application validation. Search genes by location and reliability, by tissue or cell type enrichment, or by free text and HPA search syntax. Uses HPA's keyless search API and gene pages; every result carries the release, licence and attribution.
PubChem: chemical compounds, bioassays, targets and patentsPubChem, the U.S. National Library of Medicine's open chemistry database at NCBI, through its documented PUG REST and PUG View services: find compounds by name, synonym, brand name, CAS number, SMILES, InChI, InChIKey or molecular formula; read a compound's computed properties (formula, weights, XLogP, TPSA, hydrogen-bond counts, stereochemistry), InChI, SMILES, synonyms and cross-referenced identifiers (CAS, UNII, ChEBI, ChEMBL, DrugBank, DSSTox and more); find similar compounds, substructures and superstructures ranked by Tanimoto similarity; list a compound's bioassay results across depositors and aggregate the protein and gene targets it is active against; and list the patent publications linked to it. NCBI places no restrictions on PubChem data; cross-references are identifiers only, bioassays from depositors that restrict commercial use are excluded, and every answer carries its terms and the PubChem citation.
PubMed and PMCBiomedical literature from PubMed and PubMed Central (NIH National Library of Medicine): search with PubMed query syntax, document summaries, full article records with abstracts and MeSH indexing, PMID/PMCID/DOI resolution, similar and citing articles, and open-access full text. Uses NCBI's anonymous programmatic services (E-utilities, PMC ID Converter, BioC) only; the PubMed and PMC websites themselves sit behind a proof-of-work challenge and are not scraped.
RCSB Protein Data BankExperimentally determined 3D structures of proteins, nucleic acids and their complexes from the RCSB Protein Data Bank (wwPDB archive, CC0): search by text, organism, experimental method, resolution, release date, UniProt accession or bound ligand; full entry records with polymer entities, organisms, UniProt mappings, ligands, glycans, primary citation and coordinate download links; and Chemical Component Dictionary records for ligands with identifiers and cross-references (ChEMBL, DrugBank, PubChem, ChEBI). Uses the keyless RCSB Search and Data APIs.
Reactome: biological pathways and pathway enrichmentReactome, the open, expert-curated knowledgebase of biological pathways and reactions (OICR, EMBL-EBI, NYU Langone and OHSU), through its Content Service and Analysis Service. Finds the pathways a gene, protein or chemical takes part in (gene symbol, UniProt, Ensembl, NCBI Gene or ChEBI ID), or pathways by name or topic; reads one pathway's curated summary, literature with PubMed IDs, GO process, disease terms, orthologs and its place in the pathway hierarchy; lists a pathway's participating proteins (UniProt), genes, RNAs and chemicals (ChEBI); and runs over-representation analysis of a gene list (hypergeometric test, FDR) with the genes found per pathway. Human pathways are curated and inferred for 15 model organisms; about 2,700 human pathways and 16,000 reactions. Reactome data is CC0 1.0; every answer carries the release and Reactome's citation.
NIH RePORTERNIH RePORTER research funding database via the public, keyless api.reporter.nih.gov v2 API: search NIH and other HHS-funded research projects by text, principal investigator, organization, state, fiscal year, agency and activity code; read one project with abstract, investigators, program officers, costs and study section; and list the PubMed IDs of publications linked to a project.
SEC EDGAROfficial SEC company profiles, filings, filing sections, 13F holdings, XBRL concept facts and revenue by product from each filing's XBRL, plus bulk collection of companies, XBRL facts, the daily filings index, insider transactions (Forms 3/4/5) and Schedules 13D/13G.
STRING: protein interactions and functional enrichmentSTRING (string-db.org, version 12.5; SIB, Novo Nordisk Foundation CPR and EMBL), the database of known and predicted protein-protein associations, through its public REST API (CC BY 4.0). Gives one protein's interaction partners or the network among a gene list in any of about 12,500 species, with STRING's combined confidence score and its seven evidence channels (neighborhood, gene fusion, co-occurrence, co-expression, experiments, curated databases, text mining), physical-only or functional associations, and the network's PPI enrichment p-value against chance. Runs functional enrichment of a gene list (Gene Ontology, Reactome, WikiPathways, UniProt keywords, Hallmark, COMPARTMENTS, TISSUES, DISEASES, InterPro, Pfam, PubMed, mammalian phenotypes) with FDR and the matched genes, limited to categories whose source licence permits commercial redistribution: KEGG, SMART and HPO terms are left out and counted. Names are mapped first (gene symbol, UniProt accession, Ensembl protein ID, STRING ID) and every answer names the STRING version, its licence and attribution.
UniProtUniProtKB, the UniProt Consortium's protein knowledgebase (EMBL-EBI, SIB and PIR), through its public REST API at rest.uniprot.org: search entries by text, gene, organism and Swiss-Prot review status; read one entry's protein and gene names with synonyms, organism, curated function, subcellular location with evidence codes, GO terms, disease associations, sequence length and PDB, AlphaFold, ChEMBL, Ensembl, NCBI Gene and HGNC cross-references; and map gene symbols or Ensembl, NCBI Gene, PDB and ChEMBL IDs to accessions. Data is CC BY 4.0: every result carries its license, attribution and entry URL.
More categories
Browse the library//
Get started
//
Put these providers to work.
One API key for the live web and every provider in Alexandria.
Are you an AI agent? See setup options