Drug Discovery

ADMET Prediction: In Silico Absorption,
Distribution, Metabolism, Excretion & Toxicity

Paste your SMILES strings and BioMate screens 30+ ADMET properties — hERG, CYP inhibition, BBB penetration, aqueous solubility, and toxicity flags — with Gold/Silver/Bronze QC grading and auto-remediation suggestions for failing compounds.

Try free → See drug discovery services
Property coverage

What ADMET covers

BioMate reports 30+ ADMET properties across all five pharmacokinetic and safety domains. Each property is graded against an evidence-cited threshold — not an arbitrary cutoff.

Domain Properties screened Key thresholds
Absorption Caco-2 permeability, aqueous solubility, LogP/LogD, TPSA, oral bioavailability, Pgp substrate/inhibitor Caco-2 >10 nm/s, solubility >0.1 mg/mL, TPSA <140 Ų
Distribution BBB penetration, plasma protein binding (PPB), volume of distribution (Vd), CNS MPO score BBB+ classification, PPB <99% preferred, Vd 0.1–10 L/kg
Metabolism CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP3A4 inhibition; CYP substrate prediction; metabolic stability (t½) IC50 >10 µM for non-inhibitor; t½ >30 min (microsomal)
Excretion Renal clearance, total clearance, half-life, bioavailability F% CLrenal <1 mL/min/kg preferred; t½ 4–24 h for once-daily dosing
Toxicity hERG inhibition, AMES mutagenicity, hepatotoxicity (DILI), skin sensitization, cardiotoxicity, Lipinski Ro5, Veber rules hERG IC50 >30 µM (Gold), >10 µM (Silver); AMES negative
How it works

From SMILES to QC-graded report

  1. Plain English input of compounds Paste SMILES strings, upload an SDF, or describe your compounds in plain text. BioMate parses and validates each structure automatically.
  2. Routes to ADMET workflow on AWS Batch BioMate selects the correct ADMET model ensemble, pre-fills parameters (species, target indication, safety priority), and confirms before running.
  3. Gold/Silver/Bronze per property Each of the 30+ properties receives an evidence grade: Gold (FDA/ICH consensus threshold), Silver (community standard), or Bronze (benchmark literature).
  4. Auto-remediation suggestions Failing compounds get structural modification hints or REINVENT4 generative design prompts to propose analogues that maintain potency while improving the failing property.
Example compound report
CC1=CC=C(C=C1)S(=O)(=O)N
Aqueous solubility GOLD  0.42 mg/mL
Caco-2 permeability GOLD  18.3 nm/s
hERG inhibition FAIL  IC50 6.2 µM
AMES mutagenicity GOLD  Negative
CYP3A4 inhibition SILVER  IC50 12.1 µM
hERG flag — measured 6.2 µM vs. Gold threshold 30 µM. Suggested: reduce lipophilicity or add polar substituents. REINVENT4 analogues available.
Integrated pipeline

ADMET as part of the drug discovery workflow

ADMET screening is not a standalone step — it is one gate in BioMate's end-to-end drug discovery pipeline. Results carry forward automatically to PBPK simulation and IND assembly without manual file transfers.

Target ID
→
Virtual Screening
→
ADMET Profiling
→
PBPK Simulation
→
Clinical Trial Design
→
IND Filing

Auto-remediation loop

When a compound fails an ADMET gate, BioMate's QC loop automatically suggests modified scaffolds via REINVENT4 generative design and can re-screen them in the same session.

IND §2.6 nonclinical package

ADMET results feed directly into the IND Module 2.6 DOCX. The pharmacology and toxicology sections cite each ADMET result, its QC grade, and the threshold source — ready for your RA team.

CRO submission package

ADMET reports export in CRO-ready format: compound table with all property values, failed property flags, and audit trail. Attach directly to a CRO quotation request without reformatting.

FAQ

Common questions about ADMET prediction

What input format does ADMET screening accept?

SMILES strings, SDF files, or InChI. You can paste a list of compounds directly in plain English — BioMate parses the SMILES automatically. Batch screening of up to hundreds of compounds in a single run is supported.

Which ADMET properties does BioMate report?

30+ properties covering absorption (Caco-2 permeability, aqueous solubility, lipophilicity), distribution (BBB penetration, plasma protein binding), metabolism (CYP inhibition for CYP1A2, 2C9, 2C19, 2D6, and 3A4), excretion (renal clearance, half-life), and toxicity (hERG, AMES, hepatotoxicity). Lipinski Ro5 and Veber drug-likeness rules are applied automatically.

What happens when a compound fails an ADMET threshold?

BioMate flags the failing property, shows the measured vs threshold value in Gold/Silver/Bronze format, and suggests structural modifications or analogue screening. The auto-remediation loop can automatically rescreen modified compounds via REINVENT4 generative design in the same session.

Can I run ADMET screening as part of a larger drug discovery workflow?

Yes. ADMET is one stage in BioMate's integrated drug discovery pipeline: target ID → virtual screening → ADMET → PBPK → IND filing. Results from each stage carry forward automatically, and the IND nonclinical package assembles all computational evidence into a CTD-formatted DOCX.

Get started

Screen your drug candidates in minutes

Paste your SMILES strings and get a full 30+ property ADMET report with Gold/Silver/Bronze QC grades — no chemistry informatics expertise required.

Try free →
Reading ADMET results guide → The ADMET bottleneck → All drug discovery services →