adme-assessment

name

ADME-assessment

description

Based on a compound, get its ADME and other properties. Assess compounds for ADME (Absorption, Distribution, Metabolism, Excretion) characteristics, chemical properties, and drug-likeness using PubChem. Use when the user wants to evaluate compound suitability as a lead candidate. Triggers include requests like "assess drug-likeness for [compound]", "evaluate ADME for [CID]", "ADME assessment", "check Lipinski rules for [compound]", "molecular properties of [drug]", or "is [compound] a good lead candidate". Accepts PubChem CIDs, compound names, SMILES, or InChI as input.

ADME Assessment Skill

Assess compounds for lead optimization by evaluating chemical properties, drug-likeness, and ADME characteristics via PubChem MCP tools.

Workflow Overview

1. Resolve compound identifier → Get PubChem CID from name/SMILES/InChI
2. Get molecular properties → MW, logP, TPSA, HBD, HBA, rotatable bonds
3. Assess drug-likeness → Lipinski Rule of Five, Veber rules, PAINS filters
4. Predict ADME properties → Absorption, distribution, metabolism, excretion
5. Analyze molecular complexity → Synthetic accessibility, structural features
6. Format output → Present assessment as markdown table with pass/fail indicators

Step 1: Resolve Compound Identifier

Convert user input to PubChem CID using appropriate PubChem MCP tool:

By name or identifier:

PubChem:search_compounds(query="<compound_name>")

By SMILES:

PubChem:search_by_smiles(smiles="<SMILES_string>")

By InChI:

PubChem:search_by_inchi(inchi="<InChI_string>")

By CAS number:

PubChem:search_by_cas_number(cas_number="<CAS>")

If multiple CIDs are provided (batch assessment), process each compound sequentially.

Step 2: Get Molecular Properties

PubChem:get_compound_properties(cid=<CID>)

Extract key properties:

• Molecular Weight (MW)
• LogP (partition coefficient)
• TPSA (topological polar surface area)
• HBD (hydrogen bond donors)
• HBA (hydrogen bond acceptors)
• Rotatable Bonds
• Ring Count
• Aromatic Ring Count

Also retrieve structural information:

PubChem:get_compound_info(cid=<CID>)

Returns: molecular formula, SMILES, InChI, canonical structure.

Step 3: Assess Drug-Likeness

PubChem:assess_drug_likeness(cid=<CID>)

Evaluates against established rules:

Lipinski Rule of Five

Rule	Threshold	Interpretation
MW	≤ 500 Da	Good oral absorption
LogP	≤ 5	Adequate lipophilicity
HBD	≤ 5	Hydrogen bond donors
HBA	≤ 10	Hydrogen bond acceptors

One violation allowed for oral drugs

Veber Rules (Oral Bioavailability)

Rule	Threshold
TPSA	≤ 140 Ų
Rotatable Bonds	≤ 10

PAINS Filters

Identifies pan-assay interference compounds (false positives in HTS).

Step 4: Predict ADME Properties

PubChem:predict_admet_properties(cid=<CID>)

Extract ADME-relevant predictions (exclude toxicity):

Absorption

• Intestinal absorption
• Caco-2 permeability
• P-glycoprotein substrate/inhibitor

Distribution

• Blood-brain barrier penetration
• Plasma protein binding
• Volume of distribution

Metabolism

• CYP450 substrate (CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP3A4)
• CYP450 inhibitor predictions

Excretion

• Renal clearance predictions
• Half-life estimates (if available)

Step 5: Analyze Molecular Complexity

PubChem:analyze_molecular_complexity(cid=<CID>)

Assess:

• Synthetic accessibility score (1-10, lower = easier)
• Fraction sp3 carbons (higher = more 3D character)
• Stereocenters count
• Structural complexity metrics

Optional - for pharmacophore insights:

PubChem:get_pharmacophore_features(cid=<CID>)

Step 6: Format Output

Present results in this format:

## ADME Assessment for [COMPOUND_NAME]

**Compound**: [NAME] (CID: [CID])  
**Formula**: [FORMULA]  
**SMILES**: [SMILES]

### Molecular Properties

| Property | Value | Threshold | Status |
|----------|-------|-----------|--------|
| Molecular Weight | X Da | ≤ 500 | ✓/✗ |
| LogP | X | ≤ 5 | ✓/✗ |
| TPSA | X Ų | ≤ 140 | ✓/✗ |
| HBD | X | ≤ 5 | ✓/✗ |
| HBA | X | ≤ 10 | ✓/✗ |
| Rotatable Bonds | X | ≤ 10 | ✓/✗ |

### Drug-Likeness Summary

| Rule Set | Violations | Status |
|----------|------------|--------|
| Lipinski Ro5 | X/4 | Pass/Fail |
| Veber | X/2 | Pass/Fail |
| PAINS | X alerts | Pass/Concern |

### ADME Predictions

| Property | Prediction | Confidence |
|----------|------------|------------|
| Intestinal Absorption | High/Low | X% |
| BBB Penetration | Yes/No | X% |
| CYP3A4 Substrate | Yes/No | X% |
| CYP2D6 Inhibitor | Yes/No | X% |
| ... | ... | ... |

### Complexity Analysis

| Metric | Value | Interpretation |
|--------|-------|----------------|
| Synthetic Accessibility | X/10 | Easy/Moderate/Difficult |
| Fsp3 | X | Low/Moderate/High 3D character |
| Stereocenters | X | Complexity consideration |

### Overall Assessment

[Brief summary: Is this a good lead candidate? Key strengths and liabilities.]

Error Handling

• Compound not found: If search returns no results, suggest alternative names, check SMILES validity, or ask for CID directly
• Missing ADME predictions: Note which predictions are unavailable; some compounds lack sufficient data
• Multiple matches: If search returns multiple CIDs, list top matches and ask user to specify

Example Usage

User: "Assess erlotinib as a lead compound"

Claude workflow:

1. PubChem:search_compounds(query="erlotinib") → CID 176870
2. PubChem:get_compound_properties(cid=176870)
3. PubChem:assess_drug_likeness(cid=176870)
4. PubChem:predict_admet_properties(cid=176870) (exclude toxicity)
5. PubChem:analyze_molecular_complexity(cid=176870)
6. Format and present assessment table

Batch Assessment

For multiple compounds, present a comparison table:

## ADME Comparison

| Compound | MW | LogP | TPSA | Ro5 | Veber | PAINS | Synth. Access. |
|----------|-----|------|------|-----|-------|-------|----------------|
| Compound A | X | X | X | ✓ | ✓ | ✓ | 3.2 |
| Compound B | X | X | X | ✓ | ✗ | ✓ | 5.1 |