Admetica
v1.4.1Profile small-molecule ADMET properties across 22 Chemprop-based prediction models.
Profile small-molecule ADMET properties across 22 Chemprop-based prediction models.
Admetica is Datagrok's open-source ADMET prediction toolkit for small molecules. ProteinIQ uses the published admetica==1.4.1 command-line program to predict 22 pharmacokinetic and toxicity properties from SMILES strings.
Admetica uses Chemprop graph neural networks trained per endpoint. Each model predicts one experimentally grounded ADMET property, so the tool is useful for early compound triage, lead optimization, and comparing candidates before synthesis or assay work.
Paste one SMILES per line, upload an .sdf, .csv, .smi, .smiles, or .txt file, or fetch structures from PubChem. Tab-delimited name<TAB>SMILES input is supported if you want to preserve your own identifiers.
| Format | Description |
|---|---|
| Plain SMILES | One compound per line; generated row identifiers remain unique |
| Tab-delimited | compound_name<TAB>SMILES |
| SMI | Standard SMILES name records |
| SDF | Multiple structures from a structure-data file |
| CSV | Table with a SMILES column and optional ids, id, name, compound_id, or compound identifier column |
| PubChem fetch | Resolve a compound name or CID to SMILES |
Each run supports up to 1,000 molecules. Split larger libraries into separate jobs so prediction results and the native CSV remain within the safe delivery budget.
For CSV files, SMILES column defaults to smiles; leaving it blank also uses smiles. Enter a different column name, such as Structure, to select it. Custom names are case-sensitive; the default also recognizes capitalization such as SMILES. CSV values, empty cells, and row order follow the native CSV parser. Empty SMILES cells count toward the molecule limit and retain blank predictions.
For SDF files, structures are read with RDKit, using its default removal of ordinary explicit hydrogens. Unreadable records retain a blank row and generate a warning identifying the record.
ProteinIQ supports the native property subset selector. By default, all 22 available models are selected. You can deselect endpoints you do not need to reduce result columns and run only the models relevant to your screen.
| Group | Property |
|---|---|
| Absorption | Caco2 |
| Absorption | Lipophilicity |
| Absorption | Solubility |
| Absorption | PGP-Inhibitor |
| Absorption | PGP-Substrate |
| Distribution | PPBR |
| Distribution | VDss |
| Metabolism | CYP1A2-Inhibitor |
| Metabolism | CYP2C9-Inhibitor |
| Metabolism | CYP2C19-Inhibitor |
| Metabolism | CYP2D6-Inhibitor |
| Metabolism | CYP3A4-Inhibitor |
| Metabolism | CYP1A2-Substrate |
| Metabolism | CYP2C9-Substrate |
| Metabolism | CYP2C19-Substrate |
| Metabolism | CYP2D6-Substrate |
| Metabolism | CYP3A4-Substrate |
| Excretion | CL-Hepa |
| Excretion | CL-Micro |
| Excretion | Half-Life |
| Toxicity | hERG |
| Toxicity | LD50 |
Results are returned as a spreadsheet with one row per molecule. The published runtime lowercases its prediction column keys; ProteinIQ preserves those keys, their requested order, and the source endpoint definitions.
| Column | Task | Unit or scale |
|---|---|---|
ids | Transport identifier | Submitted row |
smiles | Submitted structure | SMILES |
caco2 | Regression | log₁₀ apparent permeability in cm/s |
lipophilicity | Regression | log-ratio |
solubility | Regression | log mol/L |
pgp-inhibitor | Binary classification | Positive-class probability (0–1) |
pgp-substrate | Binary classification | Positive-class probability (0–1) |
ppbr | Regression | % bound to human plasma proteins |
vdss | Regression | L/kg |
cyp1a2-inhibitor | Binary classification | Positive-class probability (0–1) |
cyp2c9-inhibitor | Binary classification | Positive-class probability (0–1) |
cyp2c19-inhibitor | Binary classification | Positive-class probability (0–1) |
cyp2d6-inhibitor | Binary classification | Positive-class probability (0–1) |
cyp3a4-inhibitor | Binary classification | Positive-class probability (0–1) |
cyp1a2-substrate | Binary classification | Positive-class probability (0–1) |
cyp2c9-substrate | Binary classification | Positive-class probability (0–1) |
cyp2c19-substrate | Binary classification | Positive-class probability (0–1) |
cyp2d6-substrate | Binary classification | Positive-class probability (0–1) |
cyp3a4-substrate | Binary classification | Positive-class probability (0–1) |
cl-hepa | Regression | µL·min⁻¹ per 10⁶ cells |
cl-micro | Regression | mL·min⁻¹·g⁻¹ |
half-life | Regression | hr |
herg | Binary classification | Positive-class probability (0–1) |
ld50 | Regression | log(1/(mol/kg)) |
If you run only a subset of models, the output contains only ids, smiles, and the selected native property columns in the order you requested them.
The Files tab includes the exact CSV written by admetica_predict, a provenance JSON file containing the verified package version, source-distribution hash, checkpoint-bundle hash, and scientific runtime versions, and admetica_run_log.json containing complete native standard output, error output, and input warnings. The native CSV contains only the selected prediction columns; ids and smiles remain in the ProteinIQ spreadsheet for tracking.
Admetica mixes regression and classification endpoints. Continuous outputs such as caco2, solubility, cl-hepa, cl-micro, half-life, and ld50 should be interpreted in the context of the published training data and endpoint definitions. Classification outputs such as the CYP, P-gp, and herg endpoints are probabilities for the model's positive class, not measured assay values.
Invalid SMILES remain in their submitted row with blank predictions, matching the source behavior. ProteinIQ reports the affected compound identifiers in a job warning instead of silently dropping them.
Use the predictions as triage signals rather than hard acceptance rules. Compounds outside the model training domain, including unusual chemotypes and larger non-drug-like structures, can produce less reliable estimates.
ProteinIQ preserves the native predictions and the currently available 22-model selection. The tool keeps ids and smiles for workflow tracking; the native prediction CSV contains the selected model outputs.
admetica==1.4.1 runtime surface rather than every asset present in the original repository.
Predict absorption, distribution, metabolism, excretion, and toxicity properties for compound libraries.
Estimate toxicity risk and synthetic accessibility scores from SMILES for compound prioritization.
Identify toxic and reactive fragments using Brenk structural alert filter.
Screen for lead-like compounds using stricter molecular descriptor criteria for early optimization.
Screen compounds for Pan-Assay Interference patterns using PAINS_A, PAINS_B, and PAINS_C filters.
Evaluate PPI inhibitor potential and drug-likeness from molecular descriptors and SMILES inputs.
Screen compounds for structural toxicity alerts using PAINS filters, BRENK filters, and NIH filters for safer drug discovery.
Evaluate oral bioavailability from rotatable bonds, polar surface area, and molecular descriptors.
Screen compounds for oral bioavailability using molecular weight, LogP, hydrogen bond donor, and acceptor limits.
Predict protein ligand-binding residues from AlphaFold2 pair representations