Percepta: Using predictions for quick triage of legacy compounds

Description
Older internal compound collections may have limited experimental ADME/Tox data. Percepta predictions can support quick triage of which legacy compounds might be worth revisiting for new programs.

Solution

1. Assemble structures (SMILES/MOL/SDF) for a legacy compound set.
2. Run key Percepta prediction modules for each compound:
   • Physicochemical (logP/logD, solubility, pKa).
   • Selected ADME (e.g., clearance, PPB).
   • Important safety alerts or toxicity endpoints.
3. Rank or categorize compounds based on alignment with current project property targets or liability cutoffs.
4. Highlight:
   • Legacy compounds with surprisingly favorable predicted profiles that might merit reconsideration.
   • Compounds with pronounced predicted liabilities that may be poor candidates for revival.
5. Share a shortlist with project teams for further review and potential new screening or profiling.
6. As any of these compounds are re‑tested experimentally, update internal benchmarks on how Percepta predictions perform for older chemotypes.