ADME Property Predictor

Comprehensive pharmacokinetic prediction tool that assesses drug-likeness and ADME properties of small molecules using validated cheminformatics models, molecular descriptors, and structure-property relationships.

Key Capabilities:

• Multi-Property Prediction: Absorption, Distribution, Metabolism, Excretion
• Drug-Likeness Scoring: Lipinski's Rule of 5, Veber rules, QED score
• Batch Processing: Analyze compound libraries efficiently
• Structure-Based Insights: Identify liability hotspots and optimization opportunities
• Comparative Analysis: Rank candidates by predicted PK profile

Input Validation

This skill accepts: valid SMILES strings for small molecules (MW 100–800 Da), or input CSV/SMI files containing a SMILES column, plus optional property and format flags.

If the request does not involve predicting ADME properties of small molecules — for example, asking to analyze biologics, perform docking, interpret clinical PK data, or write a drug discovery report — do not proceed. Instead respond:

"ADME Property Predictor is designed for computational ADME prediction of small molecules from SMILES input. Your request appears to be outside this scope. Please provide a valid SMILES string or compound library file, or use a more appropriate tool."

Quick Check

python -m py_compile scripts/main.py
python scripts/main.py --help

Workflow

1. Confirm the SMILES input (single compound or batch file), target properties, and output format.
2. Validate that the request matches the documented scope — small molecules only; stop if the task requires unsupported assumptions.
3. Run the script or apply the documented reasoning path with only the inputs that are available.
4. Return a structured result separating assumptions, deliverables, risks, and unresolved items.
5. If execution fails or inputs are incomplete, switch to the fallback path and state exactly what blocked full completion.

Fallback: If --smiles or --input is missing, respond: "Required input not provided. Please supply a SMILES string (--smiles) or an input file (--input). Do not proceed without a valid molecular structure."

Core Capabilities

1. Absorption (A) Prediction

from scripts.adme_predictor import ADMEPredictor
predictor = ADMEPredictor()
absorption = predictor.predict_absorption(
    smiles="CC(=O)Oc1ccccc1C(=O)O",  # Aspirin
    properties=["all"]
)
print(absorption.summary())

2. Distribution (D) Prediction

3. Metabolism (M) Prediction

4. Excretion (E) Prediction

5. Drug-Likeness Scoring

CLI Usage

# Predict ADME for single compound
python scripts/main.py \
  --smiles "CC(=O)Oc1ccccc1C(=O)O" \
  --properties all \
  --output aspirin_adme.json

# Batch process compound library
python scripts/main.py \
  --input library.smi \
  --properties absorption,distribution \
  --format csv \
  --output library_adme.csv

# Filter and rank
python scripts/main.py \
  --input library_adme.csv \
  --filter "lipinski_pass=True,hia>80" \
  --rank-by qed \
  --top-n 100 \
  --output top_candidates.csv

Parameters

Output Requirements

Every final response must make these explicit:

• Objective or requested deliverable
• Inputs used (SMILES, file) and assumptions introduced
• Workflow or decision path taken
• Core result: predicted properties with values and confidence
• Constraints, risks, caveats (predictions are NOT experimental data)
• Unresolved items and next-step checks (experimental validation needed)

Error Handling

• If --smiles or --input is missing, state the missing field and request it. Do not proceed.
• If the task goes outside scope (biologics, clinical PK, docking), stop and redirect.
• If scripts/main.py fails, report the failure point, summarize what can still be completed, and provide a manual fallback.
• Do not fabricate predictions, citations, or execution outcomes.

Limitations

• Small Molecules Only: MW 100–800 Da; unreliable for biologics or macrocycles
• pH 7.4 Assumption: Most models predict at physiological pH
• Human-Specific: Predictions for human PK; animal models may differ
• Qualitative Only: For Go/No-Go decisions; not precise quantitative predictions
• No Toxicity: ADME only; use separate tools for safety assessment

Model Accuracy (Typical):

• LogP: R² = 0.85–0.95 | Solubility: R² = 0.65–0.80 | HIA: 75–85% | BBB: 70–80%

⚠️ CRITICAL DISCLAIMER: These predictions are computational estimates for prioritization only. They do NOT replace experimental ADME studies required for regulatory submissions or clinical decision-making.

References

→ Full details: references/lipinski_rules.md → Full details: references/qsar_models.md → Full details: references/property_ranges.md → Full details: references/model_validation.md

Scripts

• main.py — CLI interface
• adme_predictor.py — Core prediction engine
• absorption.py, distribution.py, metabolism.py, excretion.py — Property models
• druglikeness.py — QED, MPO scoring
• batch_processor.py — Library screening
• validator.py — Input validation and applicability domain