Scenario
A potent lead series looks great on target but may hide developability liabilities — poor absorption, rapid clearance, hERG risk, or off-target activity. Discovering these in vivo is slow and expensive. The team wants a structured early read so chemistry can design the liabilities out.
Worked example. A CNS series with strong potency must be checked for hERG and permeability risk before committing to an analog campaign.
Agent workflow
- Read the series (SMILES).
- ADMET liability triage — predict key ADMET endpoints and flag structural alerts / reactive groups.
- Selectivity & off-target screen — profile predicted off-target activity against a relevant panel.
- Aggregate into a per-compound liability scorecard.
Demo output
A liability scorecard per compound with green/amber/red flags across ADMET and selectivity, and the structural feature driving each flag. The magic moment: a top-potency compound is flagged for a reactive motif and hERG risk before it ever reaches an animal study.
Deterministic vs LLM
- Deterministic — the ADMET predictions, structural-alert matching, and selectivity scoring.
- LLM — the plain-language liability summary and the mitigation strategy.
Every flag traces to a validated model or rule. The agent reads and routes; the engine decides.
Limits
Predictive de-risking, not a replacement for measured DMPK and safety pharmacology. It focuses the assay panel; the lab confirms the profile.
Skills orchestrated by this use case
This scenario routes an AI agent through the following curated skills.
-
ADMET Liability Triage
Medicinal ChemistryTriage a set of candidate molecules for early ADMET/developability liabilities by orchestrating structure standardization, physicochemical descriptors, rule-based flags (Lipinski/Veber/PAINS/reactive), and ML ADMET predictions, then ranking into keep/review/drop tiers.When to use: a user hands over SMILES/a library and asks which compounds to prioritize.Multi-tool WORKFLOW, not one call. -
Selectivity Off-Target Screen
Medicinal ChemistryScreen candidate molecules for selectivity and off-target risk by predicting activity against anti-targets (hERG, CYPs, key receptors) and profiling similarity to known liabilities using Paramus BRAIN models, then flagging selectivity concerns.When to use: a user asks “is this compound selective / what might it hit off-target?”.Multi-tool WORKFLOW with safety-pharmacology judgment.
Tools it reaches for
- ADMET property prediction (absorption, clearance, hERG, etc.)
- Structural-alert and reactive-group screening
- Off-target / selectivity profiling
Browse the full deterministic layer in the tool browser.