Search any human gene for its structure, constraint, protein, disease associations and
genome-browser links.
Try:
Traxome Graph
◈
Traxome Graph
A connected map of the biomedical world — genes, conditions, drugs and pathways, wired together with multi-hop discoveries. Search an entity to begin, or start from one of these:
Methods & Glossary
Traxome assigns every variant a classification with an explicit, auditable rule set. There is no
learned model and no hidden weighting: the verdict is an integer sum of named criteria, each traceable to a
public database. This page documents the model in full so any result can be reproduced by hand — and the
glossary below defines exactly what each term means in this platform.
1 · Additive evidence model
Each variant accrues an evidence score \(S\) as the signed sum of the contributions \(w_i\) of the criteria
that fire for it. Positive weights push toward pathogenic; negative weights toward benign.
where \(\mathcal{C}\) is the set of criteria evaluated for the variant and \(\mathbb{1}[\cdot]\)
is the indicator that criterion \(i\) fired given the observed evidence \(e\).
2 · Tier thresholds
The integer score \(S\) maps to a five-tier classification by fixed thresholds:
3 · Criteria and weights
Every criterion, its ACMG/AMP-style code, the evidence that triggers it, and its integer weight:
Code
Criterion
Weight
4 · Data sources
Every value shown in a result is a live lookup — nothing is cached or hand-curated. Sources and versions:
Layer
Source
5 · Scope and limits
This additive model is inspired by the ACMG/AMP sequence-variant interpretation guidelines
(Richards et al., Genet Med 2015) but is deliberately simplified for transparency and speed. It is a
decision-support aid, not a certified clinical classification. A qualified clinical geneticist must confirm any
interpretation before clinical use.
6 · Glossary — what these words mean here
Terms are used with specific meanings in this platform. This is what each one denotes on a Traxome result.
Signature section
The most important, decision-driving part of a dossier — the medications for a condition, the FDA label for a drug, the risk genes for a researcher, the 3D structure for a gene. Traxome marks it with a Signature badge and opens it by default so the thing that matters most is never buried.
Verdict tier
One of five classifications — Pathogenic, Likely pathogenic, Uncertain significance (VUS), Likely benign, Benign — assigned purely by the integer evidence score above. Not a diagnosis.
Evidence score
The signed integer sum of the ACMG-style criteria that fired for a variant. Positive → pathogenic, negative → benign. Fully reproducible from the weights table.
Actionable (pharmacogenomics)
A gene–drug pair where a person's genotype should change prescribing — i.e. CPIC evidence level A or B. "Informational" means a relationship exists but no prescribing change is recommended (CPIC level C/D).
CPIC level
The Clinical Pharmacogenetics Implementation Consortium's evidence grade for a gene–drug pair. A/B = actionable, C/D = informational.
pLI
gnomAD probability a gene is intolerant of loss-of-function variation (0–1). pLI > 0.9 flags a likely haploinsufficient gene.
LOEUF
Loss-of-function observed/expected upper-bound fraction (gnomAD). Lower = more constrained; LOEUF < 0.35 marks a strongly LoF-intolerant gene. Batch gene panels are ranked by LOEUF.
Missense-Z
gnomAD constraint on missense variation; Z > 3.09 indicates a gene depleted of missense variants.
Association score
Open Targets' evidence-weighted disease↔gene association (0–1), aggregating genetics, literature, clinical and animal evidence. Drives the researcher risk-gene ranking.
Architecture (risk genes)
How a gene is linked to a disease: Mendelian / rare (rare-disease genetics evidence), GWAS / common (common-variant association), Clinical evidence, or Drug target (target of a known drug).
AlphaMissense · PolyPhen-2 · SIFT · ClinPred
In-silico predictors of whether a missense change is damaging. Traxome shows each one's call and links its source; it never lets a predictor alone decide the verdict.
pLDDT
AlphaFold's per-residue model confidence (0–100), shown as the colour of the 3D structure. Blue/high = confident, orange/low = uncertain.
FAERS
The FDA Adverse Event Reporting System — real-world reports of events that occurred while people took a substance. A signal, not proof of cause: reports include coincidental events and reasons for use.
Investigational
A drug shown for a condition that is not FDA-approved for it — a research or clinical-trial candidate, not an approved treatment.
Substance
Any drug, supplement, herb, peptide, nootropic or vitamin. Supplements are not FDA-reviewed for safety/efficacy and product quality varies.
Active constituent
A specific compound responsible for a substance's activity. A brand or herb is shown by its active ingredient(s) — e.g. Advil by ibuprofen, ashwagandha by its withanolides.
About Traxome™
Traxome is a transparent interpreter for human biology. Type a condition, gene, variant,
medicine or supplement and get a clear, sourced answer.
The Patient · Clinician · Researcher lens doesn't just hide detail — it fetches genuinely
different information from the sources that matter to each reader.
Patient — plain-language explainers, side effects and interactions in everyday words, trusted resources (MedlinePlus/NIH, Mayo, Cleveland, NINDS, NHS), and where to find or afford a treatment.
Clinician — FDA-label detail (uses, interactions, contraindications, adverse reactions), real-world safety signal (FAERS), pharmacogenomics (CPIC), and guideline links.
Researcher — evidence-ranked risk genes, genetic architecture, chemistry and mechanism, plus a Gene Explorer with an interactive 3D protein structure.
What you can ask
Conditions (migraine), genes (CACNA1A), variants (BRCA1 R1699W), pharmacogene pairs
(CYP2C19 clopidogrel), and anything you can take — prescription drugs, supplements, herbs, peptides and
nootropics (advil, ashwagandha) — each identified and given the right information and where to
research or buy it.
For developers — Evidence API & MCP
Traxome's deterministic evidence is callable by any tool through one registry of read-only tools, exposed as
both a REST API and an MCP server — so an AI agent, notebook, or lab pipeline can consume it as
data (with provenance and a re-verifiable capsule hash), not scraped HTML. Both are off by default; enable with
TRAXOME_EVIDENCE_API=1.
# REST — list the tools, then call one
TRAXOME_EVIDENCE_API=1 ./run.sh
curl localhost:8069/api/v1/tools
curl -X POST localhost:8069/api/v1/tools/get_variant_evidence \
-H 'content-type: application/json' -d '{"params":{"query":"BRCA1 R1699W"}}'
# MCP (local, stdio) — the same tools; plug into Claude Desktop, Cursor, any MCP client
TRAXOME_EVIDENCE_API=1 python -m evidence_api.mcp
# MCP (remote) — connect by URL in Claude → Settings → Connectors → Add custom connector
https://traxome.vercel.app/mcp # leave OAuth Client ID blank (auto-approve, no login)
Tools: resolve_entity, get_gene_evidence, get_variant_evidence,
get_gene_disease_links, get_pharmacogenomic_guidance, get_controlled_status,
search_literature, build_capsule, validate_capsule, compare_capsules. Responses carry
provenance + version; errors are stable typed codes; raw notes / VCFs / PHI are refused.
Data sources
ClinVar · gnomAD · dbNSFP (AlphaMissense, PolyPhen-2, SIFT, ClinPred, GERP) · UniProt · AlphaFold · RCSB/PDBe ·
Ensembl · Open Targets · CPIC · openFDA (labels + FAERS) · RxNorm · PubChem · HPO / Monarch · MONDO ·
PubMed · Europe PMC · Crossref · arXiv · ClinicalTrials.gov · MedlinePlus (NIH). All retrieved live at query time.
Full list of databases & sources61 sources · fully documented
Everything Traxome cites is public, free, and open. Most sources below are queried live and keyless at the moment you search; a few registries and patient-reference sites are linked for you to read directly. Every link opens in a new tab. This is the complete list — if a source appears anywhere in a Traxome answer, it is documented here.
A public archive of reported human genetic variants and their classified clinical significance for diseases and drug responses, with supporting evidence.
How it's collected
Aggregates voluntary submissions from clinical labs, research groups, and expert panels, organizing them into variant (VCV) and variant-condition (RCV) records.
Update frequency
Weekly (records refreshed on a roughly weekly release cycle, with full monthly data dumps).
Why it's credible
Maintained by NCBI at the NIH National Library of Medicine, the standard reference archive for variant interpretations.
Limitations
Submitted classifications can conflict between labs and vary in evidence quality, so a single ClinVar record is not a definitive clinical determination.
A reference database of population allele frequencies aggregated from large-scale exome and genome sequencing of hundreds of thousands of individuals.
How it's collected
Harmonizes and jointly reprocesses exome and genome data contributed by an international coalition of sequencing projects, filtering to unrelated individuals.
Update frequency
Roughly annual major/minor releases (current v4.1 on GRCh38); no fixed calendar cadence.
Why it's credible
Produced by the Broad Institute with a global investigator coalition, the field-standard source for population frequencies.
Limitations
Ancestry representation is uneven (skewed toward European samples) and it deliberately removes known disease cohorts, so frequencies are not fully representative of all populations.
A precomputed database of functional annotations and deleteriousness prediction scores for all possible human nonsynonymous and splice-site SNVs.
How it's collected
Compiles and aligns scores from 36+ prediction algorithms and conservation metrics onto every possible coding SNV in the human genome.
Update frequency
Periodic versioned releases tied to new GENCODE/Ensembl builds (latest release January 2026); no fixed schedule.
Why it's credible
Long-standing academic resource (originated by the Liu lab) widely used as the aggregation layer for variant-effect predictors, now maintained by Genos Bioinformatics.
Limitations
It only redistributes third-party predictor scores, which are computational estimates that often disagree with each other and are not clinical evidence.
NCBI's public catalog of short genetic variations (SNVs, small indels, microsatellites) with stable reference SNP (rsID) identifiers.
How it's collected
Aggregated from community and project submissions, then clustered into non-redundant reference SNP (rs) records mapped to reference assemblies with population-frequency and functional annotation.
Update frequency
Released in periodic numbered builds; no fixed public schedule.
Why it's credible
Operated by NCBI (National Library of Medicine, NIH) as a long-standing authoritative public genomic-variation resource and the canonical source of rsIDs.
Limitations
Being submission-driven it includes variants of varying validation and quality and is a variant catalog, not a clinical-significance authority; Traxome uses it for rsID lookup, not for interpreting pathogenicity.
An open database of AI-predicted 3D protein structures covering over 200 million proteins.
How it's collected
Structures are computationally predicted by DeepMind's AlphaFold model from UniProt amino-acid sequences, not experimentally determined.
Update frequency
Updated irregularly in periodic major releases as models improve and new sequences/features are added.
Why it's credible
Built by Google DeepMind and hosted by EMBL-EBI, with AlphaFold's accuracy validated in the CASP assessments.
Limitations
All entries are predictions with per-residue confidence scores, and low-confidence or disordered regions can be unreliable; not a substitute for experimental structures.
The US portal to the global archive of experimentally determined 3D biomolecular structures.
How it's collected
Researchers deposit structures through the wwPDB OneDep system, where they are validated and biocurated before release.
Update frequency
New and updated entries are released weekly.
Why it's credible
A wwPDB founding member funded by NSF, DOE, and NIH and hosted by Rutgers, UC San Diego, and UCSF.
Limitations
Structures reflect experimental conditions and vary in resolution/quality; coverage is limited to proteins that have been successfully crystallized or imaged.
The European node of the worldwide PDB, providing the same structure archive plus added analysis services.
How it's collected
Shares the single wwPDB deposition and curation pipeline, adding value services like SIFTS mappings, PDBeFold, and PDBePISA.
Update frequency
Synchronized with the wwPDB weekly release cycle.
Why it's credible
Operated by EMBL-EBI as a founding wwPDB partner alongside RCSB PDB and PDBj.
Limitations
Draws on the same underlying archive as RCSB, so it shares PDB's experimental-quality and coverage limits; its extra tooling has a steeper learning curve.
An integrative human gene database compiling one searchable card per gene from many sources.
How it's collected
Automatically mines and integrates information from more than 190 web sources into a unified per-gene summary.
Update frequency
Updated in periodic versioned releases (no fixed public schedule), with continuous underlying data refreshes.
Why it's credible
Developed by the Weizmann Institute of Science with LifeMap Sciences, a long-established gene compendium.
Limitations
Free only for non-profit academic and clinical use (commercial license otherwise); as an automated aggregator it can surface inconsistent or outdated entries from its sources.
An open-source, web-based modular genome browser for visualizing genomic data (alignments, variants, annotations, synteny) at a locus.
How it's collected
It is visualization software, not a dataset; it renders whatever genome assemblies and track files (BAM/VCF/GFF, etc.) are loaded into it rather than curating its own data.
Update frequency
Released continuously as versioned software on an ongoing open-source development cadence; no fixed schedule.
Why it's credible
Developed by the JBrowse Consortium under the Apache 2.0 license with NIH, Chan Zuckerberg Initiative, OICR and UC Berkeley funding, and a peer-reviewed reference (Diesh et al., Genome Biology 2023).
Limitations
A rendering tool, not a primary data source, so its accuracy depends entirely on the underlying tracks; Traxome only deep-links to it at a gene locus and does not treat it as evidence.
A freely available platform that integrates genetics, genomics, and other evidence to score and prioritize therapeutic drug targets and their disease associations.
How it's collected
Systematically ingests and harmonizes many public datasets (e.g. GWAS, ChEMBL, literature, expression) into standardized target-disease evidence and computed association scores.
Update frequency
Quarterly data releases (roughly four per year), versioned as year.month (e.g. 26.06).
Why it's credible
A public-private partnership led by EMBL-EBI, the Wellcome Sanger Institute, and pharmaceutical partners.
Limitations
Association scores are algorithmic and evidence-weighted, so they reflect data availability and heuristics rather than confirmed biological causation.
An international consortium that publishes peer-reviewed, freely available clinical guidelines translating specific gene-drug pairs into actionable prescribing recommendations.
How it's collected
Expert author panels systematically review the literature and grade evidence to reach consensus guideline recommendations and assign standardized allele-function and phenotype terms.
Update frequency
New and revised guidelines are published on a rolling per-drug basis (no fixed calendar), with periodic supplement and database/API updates.
Why it's credible
Co-led by Stanford and St. Jude, jointly funded by NIH, with guidelines published in peer-reviewed journals and endorsed for clinical use.
Limitations
Coverage is limited to gene-drug pairs the consortium has chosen to curate, so many drugs and genes have no CPIC guideline; the standalone cpicpgx.org site now redirects to ClinPGx and is being retired.
A comprehensive pharmacogenomics knowledge base of gene-drug-variant relationships, clinical annotations, drug pathways, and dosing guideline summaries, now hosted on the ClinPGx platform.
How it's collected
A scientific curation team manually reviews and annotates published pharmacogenomic literature, assigning levels of evidence and linking variants, drugs, and phenotypes.
Update frequency
Continuously curated with periodic data releases; content is updated on an ongoing rolling basis rather than a fixed schedule.
Why it's credible
Long-running NIH/NHGRI-funded resource based at Stanford University and widely cited as the primary reference for pharmacogenomic annotations.
Limitations
Annotations reflect published evidence of variable strength, so many associations are preliminary; as of July 2025 pharmgkb.org permanently redirects to clinpgx.org, so older PharmGKB links now resolve to the merged site.
The authoritative catalog and standardized star-allele nomenclature for pharmacogenes such as the cytochrome P450 (CYP) family.
How it's collected
International gene-specific expert panels curate reported allelic variation and define standardized star-allele haplotypes with defining sequence variants.
Update frequency
Versioned releases issued periodically as genes and alleles are added or revised (e.g., recent versions added NAT2 and CYP1A2), without a fixed calendar.
Why it's credible
The successor to the Human CYP Allele Nomenclature Database and the nomenclature source relied on by CPIC and PharmGKB, curated by recognized pharmacogenetics experts.
Limitations
Scope is confined to a defined set of pharmacogenes, and evolving allele definitions can change diplotype and phenotype calls between versions.
The FDA's open-data platform providing public APIs and downloads for structured drug product labeling, adverse-event reports (FAERS), enforcement/recall, and related datasets.
How it's collected
It aggregates and normalizes existing FDA regulatory submissions and reports into queryable JSON APIs and bulk downloads.
Update frequency
Structured drug labeling is refreshed roughly weekly while FAERS adverse-event data is loaded quarterly, so different endpoints have different cadences.
Why it's credible
Operated directly by the U.S. Food and Drug Administration as its official open-data service.
Limitations
FAERS reports are unverified, spontaneous, and cannot establish causation, and the API is explicitly not intended for clinical decision-making or as the authoritative label of record.
The NLM's searchable database of official FDA-approved drug labeling (package inserts / structured product labels) for medications marketed in the US.
How it's collected
It publishes the SPL labeling documents submitted by manufacturers to the FDA, presenting the current in-use label without editorial modification of content.
Update frequency
Updated daily as new and revised labeling submissions are received from the FDA, with an RSS feed for changes.
Why it's credible
Maintained by the U.S. National Library of Medicine (NIH) as the official public source of FDA drug label content.
Limitations
It reflects labels as submitted by manufacturers, so information can lag real-world evidence and label wording varies in completeness across products.
RxNorm is a standardized normalized naming system for clinical drugs, and RxNav is the browser/API interface for exploring it and linked drug vocabularies.
How it's collected
NLM curators integrate and cross-map drug names from multiple source vocabularies into normalized concepts with unique RxCUI identifiers.
Update frequency
A full monthly release on the first Monday of each month plus weekly incremental updates every Wednesday for newly approved drugs.
Why it's credible
Produced and maintained by the U.S. National Library of Medicine (NIH) as a core standard for US drug interoperability.
Limitations
It normalizes drug naming and identifiers rather than providing clinical guidance, and the live RxNav interface depends on the RxNorm API, which can be transiently unavailable.
A very large open chemistry database of compounds, substances, bioassays, and associated literature and safety information.
How it's collected
It aggregates depositor-contributed substance records and hundreds of external data sources, automatically standardizing them into unique compound records.
Update frequency
Continuously updated as depositors and integrated sources contribute data, with regular ongoing ingestion rather than a single fixed release date.
Why it's credible
Maintained by NCBI at the U.S. National Library of Medicine (NIH), a primary public cheminformatics resource.
Limitations
Data quality varies by depositor and source, so records can contain inconsistencies, duplicates, or unvalidated bioactivity annotations.
A detailed drug and drug-target knowledge base combining chemical, pharmacological, pharmacokinetic, and target/interaction data for approved and investigational drugs.
How it's collected
A dedicated team of curators manually extracts and reviews data from primary literature, regulatory labels, and other sources, supplemented by external database integration.
Update frequency
Data is curated on an ongoing basis with incremental versioned exports, alongside larger periodic major releases.
Why it's credible
Developed at the University of Alberta in a public-private partnership with OMx Personal Health Analytics and widely cited in pharmacology research.
Limitations
Free access is restricted to non-commercial use under a CC BY-NC license, and the homepage blocks automated fetchers (returns HTTP 403 to bots) though it loads normally in a browser.
A freely available reference database of small-molecule human metabolites with their chemical, biological, and clinical/physiological data.
How it's collected
Curators compile metabolite data from literature, experimental NMR/MS spectra, and other databases, combining manual annotation with computational prediction.
Update frequency
Released as major numbered versions (current major release 5.0, published 2022) with interim additions between major versions.
Why it's credible
Maintained by researchers at the University of Alberta and a foundational resource for metabolomics, described in peer-reviewed Nucleic Acids Research papers.
Limitations
Some metabolite concentrations and annotations are predicted or drawn from limited studies, and the homepage returns HTTP 403 to automated fetchers though it loads normally in a browser.
The DEA division that regulates controlled substances and publishes the official US controlled-substance schedules, registration rules, and related regulatory notices.
How it's collected
It reflects the legally binding schedules and amendments established through the Controlled Substances Act, federal rulemaking, and Federal Register notices.
Update frequency
Updated as scheduling actions and regulations are issued (event-driven via Federal Register), with no fixed periodic cycle.
Why it's credible
Operated by the U.S. Drug Enforcement Administration (Department of Justice) as the authoritative source for federal controlled-substance status.
Limitations
It covers US federal scheduling only and does not reflect differing state schedules or international controls, and legal text can be dense and time-lagged from enacted changes.
The NIH office providing evidence-based fact sheets and research databases on dietary supplements and their ingredients for consumers and professionals.
How it's collected
ODS staff synthesize peer-reviewed scientific evidence into fact sheets and maintain structured databases such as the Dietary Supplement Label Database (DSLD) and CARDS.
Update frequency
Fact sheets are revised periodically as evidence changes (each carries its own last-updated date) rather than on a fixed schedule.
Why it's credible
Part of the U.S. National Institutes of Health, offering government-authored, non-commercial evidence reviews.
Limitations
It provides general educational information rather than individualized medical advice, and supplement evidence is often limited or evolving so guidance can be cautious or incomplete.
A large, manually curated database of bioactive drug-like molecules with their bioactivity, mechanism-of-action, and target information.
How it's collected
Compiled by expert manual curation and extraction from the primary medicinal-chemistry literature plus deposited datasets, mapped to standardized targets and assays.
Update frequency
Published in periodic numbered releases (roughly one to a few per year); no fixed calendar cadence.
Why it's credible
Maintained by EMBL-EBI as an ELIXIR Core Data Resource / Global Core Biodata Resource, released under CC BY-SA 3.0, and widely cited in drug-discovery research.
Limitations
Bioactivity values are aggregated from heterogeneous assays of varying quality and standardization, and coverage lags the newest literature; Traxome queries it live for MoA/bioactivity but values should be read in assay context.
Free biomedical literature database of 40M+ citations from MEDLINE, life-science journals, and online books, with programmatic access via the NCBI E-utilities API.
How it's collected
Curated by indexing peer-reviewed journal citations into MEDLINE and ingesting publisher-supplied and PMC records under NLM editorial standards.
Update frequency
Daily additions to the live database, with an annual baseline release each January (2026 baseline released Jan 30, 2026).
Why it's credible
Operated by the U.S. National Library of Medicine at the NIH, the reference index for biomedical literature.
Limitations
Coverage is citation/abstract-level (not full text) and skews toward MEDLINE-indexed journals, so many preprints and non-indexed sources are absent.
Free search interface covering 120M+ patent publications and scholarly documents from patent offices worldwide.
How it's collected
Aggregates bibliographic data, full text, and machine translations from national/regional patent offices, sourced largely via IFI CLAIMS.
Update frequency
Underlying public dataset targets quarterly refreshes, though actual update timing can lag.
Why it's credible
Operated by Google using licensed patent data from IFI CLAIMS and official patent authorities.
Limitations
Not an official legal register; classifications, machine translations, and status/legal-event data can be inaccurate or stale, so filings should be verified against source offices.
A persistent-identifier system (ISO 26324) that assigns stable, resolvable names to publications and other objects, enabling durable links via doi.org.
How it's collected
DOI names and their metadata are registered by Registration Agencies (e.g., Crossref, DataCite) and resolved through shared Handle-system infrastructure rather than being a content dataset.
Update frequency
Resolution service runs continuously in real time; registrations are added on an ongoing basis by agencies with no fixed cadence.
Why it's credible
Governed by the not-for-profit DOI Foundation as the ISO 26324 registration authority, with technical infrastructure operated by CNRI and billions of resolutions served.
Limitations
It is an identifier/resolution layer, not a source of paper content or credibility, and a resolved DOI only guarantees identity/location, not open access to the target; Traxome uses it purely to build stable links.
Free consumer-health information portal covering conditions, drugs, supplements, genetics, and a medical encyclopedia.
How it's collected
Written and curated by NLM staff drawing on NIH institutes and other vetted authorities, with an advertising-free editorial policy.
Update frequency
Continuous: health-topic pages updated as needed and links checked daily; encyclopedia and drug pages refreshed monthly; content fully reviewed at least every 3 years.
Why it's credible
Produced by the US National Library of Medicine at NIH, the world's largest biomedical library.
Limitations
Written for general patient audiences, so it lacks the depth, primary citations, and specialist nuance clinicians may need.
Academic medical center's site offering patient-facing health information, conditions, and treatment overviews.
How it's collected
Health-library content is authored and reviewed by Johns Hopkins faculty and clinical staff, some licensed from vetted third-party health-content providers.
Update frequency
Periodic and inconsistent: pages are revised as needed with no single published update cadence across the library.
Why it's credible
Produced by Johns Hopkins Medicine, a world-renowned academic and research hospital.
Limitations
Some library content is syndicated from external vendors and lightly dated, and articles are general rather than exhaustively referenced.
Official health information and service-access site for England, covering conditions, symptoms, medicines, and treatments.
How it's collected
Content is produced and clinically assured by NHS editorial teams following an evidence-based information standard.
Update frequency
Per-page review cycle: most condition pages display a last-reviewed date and a next-review date, typically about every 3 years, with interim edits as needed.
Why it's credible
Official England-wide public health service, published under Crown copyright.
Limitations
Reflects UK/NHS clinical practice, drug names, and service pathways that may not match care or terminology in other countries.
UK body publishing evidence-based clinical guidance, quality standards, and technology/drug appraisals for the NHS.
How it's collected
Guidance is developed by independent committees through systematic evidence review, expert input, and public consultation.
Update frequency
Rolling surveillance model: guidelines are monitored continuously and updated (including modular updates) when new evidence warrants, so intervals vary by topic.
Why it's credible
Statutory UK public body whose recommendations are the reference standard for NHS practice.
Limitations
Recommendations are calibrated to the UK health system and cost-effectiveness thresholds, and appraisals can lag the newest evidence or therapies.
Patient-advocacy nonprofit hosting a rare-disease database with condition reports plus patient-assistance resources.
How it's collected
Rare Disease Database reports are authored or reviewed by medical experts, often in partnership with specialist physicians and academic contributors.
Update frequency
Periodic, per-report: individual disease entries carry a last-updated year and are revised as new information and expert review allow, not continuously.
Why it's credible
Established US rare-disease advocacy organization working with clinical and academic experts.
Limitations
Some entries are infrequently updated given the pace of rare-disease research, and depth varies by how well-studied a condition is.
A subscription point-of-care clinical decision-support reference of evidence-based, expert-authored clinical topic reviews.
How it's collected
Written and continuously revised by physician authors, editors, and peer reviewers who synthesize the medical literature into graded clinical recommendations.
Update frequency
Continuously updated as evidence and editorial review warrant; not on a fixed public schedule.
Why it's credible
Published by Wolters Kluwer with a large physician editorial process and broad clinical adoption, and widely used at the point of care.
Limitations
Proprietary and paywalled behind an institutional/individual subscription, so full content is not openly accessible or programmatically reusable; Traxome only links out to it.
A subscription evidence-based clinical reference providing point-of-care summaries and graded recommendations for clinicians.
How it's collected
Physician and evidence-team editors systematically monitor and appraise the literature, then distill it into structured topic summaries with explicit evidence grading.
Update frequency
Updated on a daily/continuous basis as new evidence is identified; no single fixed release date.
Why it's credible
Produced by EBSCO with a documented systematic evidence-surveillance and expert-review methodology, and established clinical adoption.
Limitations
Proprietary and subscription-gated, so full content requires paid access and is not openly reusable; Traxome only links out to it.
A general-reference fallback link, offered only when a term has no structured dossier in the biomedical databases above.
How it's collected
Openly, community-written and community-edited by volunteers, with citations to external sources.
Update frequency
Continuous (edited in real time).
Why it's credible
Broad coverage with a citation culture; useful as a starting point, not an authority.
Limitations
Anyone can edit it, so accuracy varies by article — never treat it as a primary or clinical source.
Price & availability links
For non-controlled medicines and supplements, Traxome may show external "where to find or afford it" links — e.g. GoodRx, Cost Plus Drugs, TrumpRx (a US federal drug-price finder), and major retailers (Amazon, Walmart, Target, iHerb).
What these are
Convenience search links to third-party pharmacies and retailers — not evidence sources, and not endorsements.
Limitations
Prices, stock and sellers are set by those third parties; Traxome does not verify them and never links purchase pages for controlled substances.
