Last updated: April 28, 2026 · Liganx Beta
Liganx is a research-preview tool that runs computational molecular docking (AutoDock Vina + Vinardo rescoring) against wild-type and mutant protein structures, with mutation models built by FoldX or PDBFixer and contact analysis by ProLIF + PoseBusters. It exists to help you triage compounds and mutations quickly, not to replace wet-lab validation or clinical decision-making.
Docking scores are not binding affinity measurements. They're a numerical heuristic for ranking poses and (with care) comparing close analogs. Vina scores at exhaustiveness 8 have typical noise of ±1 kcal/mol; deltas under that are unreliable. Single-conformation docking can't capture allosteric effects, induced fit, or the consequences of mutations that change protein conformation rather than pocket geometry. Cells with a "outside pocket" badge mean exactly that — the method can't see those mutations and is honestly telling you so.
Liganx output is not medical advice and not a substitute for clinical, regulatory, or experimental validation. Don't make patient-care decisions from a Liganx matrix.
You may use Liganx for academic research, drug-discovery exploration, teaching, and personal learning. Don't submit structures or compounds you don't have rights to share, don't try to use the API as a public file host, and don't scrape the public job pages (the share IDs are random per job, but mass-enumeration is still abuse). Reasonable rate limits apply on a per-IP basis to keep the GPU backend responsive for everyone.
Liganx is provided as-is during the beta period without warranty. We try to keep the API up but don't guarantee uptime. Catalog data (PDB structures, mutation libraries, reference compounds) is curated from public sources and may contain errors — see the matrix provenance tags (foldx_precached / pdbfixer_mutated / mutant_verify_failed / mutation_outside_pocket) for the per-cell confidence story. If you spot an error, please tell us.
To the extent permitted by law, Liganx and its operators are not liable for any direct, indirect, incidental, or consequential damages arising from use of the platform, including but not limited to lost research time, missed lab opportunities, or decisions made on the basis of computational predictions.
We update these terms occasionally. Material changes will be flagged on the homepage with a banner for at least a week before they take effect. See also our Privacy Policy.