Versioned data release
Download the peptide ADMET evidence release.
One release, three ways in: take the complete release, an analysis-ready subset scoped to an ADMET domain or evidence tier, or the versioned API. Every measurement keeps its unit, assay context, and source citation so the values stay interpretable downstream.
- Release
- ADMETatlas V1.0
- Last updated
- 2026-08-11
- Identity records
- 37,467
- Measurements
- 39,641
- Endpoints
- 17
- Products
- 110
These counts show where evidence exists. They do not rank peptide safety, efficacy, or developability by themselves.
Endpoint values require unit or metric scale, model, matrix, route, species, concentration, and method fields before comparison.
Trials, labels, adverse-event text, and immune-recognition entries remain contextual unless review policy promotes them.
Complete release
Every table in one release
Each entry below is a table with its current record count. Take the whole bundle, or a single table in TSV — sequence tables also export FASTA. Verify any download against the published SHA-256 checksums.
| Table | Contents | Rows | Download |
|---|---|---|---|
Canonical identity records peptides.tsv | Canonical identity records at the resolution supported by each source. Sequence availability is reported separately by peptides.fasta. | 37,467 | |
Peptidoforms peptidoforms.tsv | Peptide forms: sequence/form identifier, topology, and modification summary. | 40,948 | |
Endpoint definitions endpoints.tsv | Formal ADMET endpoint definitions with default unit, minimum context, and per-endpoint counts. | 17 | |
Endpoint measurements evidence_measurements.tsv | Reference-linked measurements with value, unit, inline assay context, and per-measurement source citation. | 39,641 | |
Endpoint summaries endpoint_summaries.tsv | Derived quantitative summaries stratified by conditions recorded in the public measurement evidence, with the supporting public measurement identifiers for each summary. | 35,719 | |
Assay contexts assay_contexts.tsv | Distinct public assay contexts referenced by release measurements, including species, matrix, route, method, and timing fields when reported. | 3,888 | |
Context signals context_signals.tsv | Trial, label, safety, and immune-recognition context bound to a peptide or product. | 3,678 | |
Drug products products.tsv | Drug-product identity, route/form, status, and its linked peptide identity. | 110 | |
Product–peptide links product_peptide_bridge.tsv | Product–peptide links with their link status and identity-support role. | 115 | |
Computed descriptors computed_descriptors.tsv | Deterministic molecular descriptor values, kept separate from measured endpoint evidence. | — | |
Source records provenance.tsv | Public source-record identifiers, citations, and traceability fields referenced by release measurements. | 43,276 | |
Source collections source_manifest.tsv | Collection-level source, attribution, license, and redistribution metadata for this release. | — |
Assay conditions also travel inline with each measurement. The separate context table provides stable context identifiers for joins and release-level reuse.
Analysis-ready subsets
Pre-filtered for a specific question
Skip the join. Each subset is a single download already filtered along one scientific axis — the ADMET domain, the evidence tier, or the molecule — with the same provenance columns as the full release.
By ADMET domain
Endpoint measurements, one domain at a time
Every measurement for the endpoints in a domain, with assay context and source citation inline.
Exposure / Absorption
Bioavailability, AUC, Cmax, and Tmax evidence.
Distribution / Barrier
Distribution volume, plasma protein binding, permeability, and BBB penetration evidence.
Metabolic & Proteolytic Stability
Serum/plasma stability and protease stability evidence.
Excretion / Persistence
Clearance and half-life evidence.
By evidence quality
The governance axis, as a download
Each tier carries its scientific definition. Comparable measurements, reviewed evidence, literature-traceable measurements, and context are kept in separate files.
Curated comparable measurement
Structured measurements that can be compared within their endpoint and assay context.
Reviewed reported evidence
Reviewed reported evidence not yet treated as a comparable measurement.
Reference-linked evidence
Measurements carrying a PMID, DOI, or source URL for literature tracing.
Context only
Clinical, product, identity, and provenance context kept separate from endpoint measurements.
By molecule & product
Sequences and the product layer
Peptide identities as sequences, and the drug-product layer that points back to them.
Canonical identity records
The TSV retains all canonical identity records, including source-scoped unresolved identities. The FASTA contains 22,605 records with a non-empty plain sequence.
Peptidoforms
Peptide forms with topology class and modification summary.
Programmatic access
The same release, as a versioned API
Every table and subset is also a versioned HTTP endpoint. Use the API for scripted retrieval, server-side filtering, or exact release capture.
curl -sS 'https://admetatlas.scbdd.com/api/v1/downloads/release_manifest.json' \
-H 'accept: application/json' \
-H 'X-Visibility: public_release'