Protein Post-Translational Modification Analysis

Comprehensive PTM analysis using iPTMnet (primary), ProtVar (functional context), UniProt (baseline), STRING (interactions), ELM (linear motifs), and MassIVE/ProteomeXchange (experimental data).

LOOK UP DON'T GUESS

• PTM sites/enzymes: iPTMnetgetptm_sites
• Functional consequence: ProtVargetfunction + iPTMnetgetptm_ppi
• Proteoforms: iPTMnetgetproteoforms
• Linear motifs: ELMgetinstances

COMPUTE, DON'T DESCRIBE

When analysis requires computation (statistics, data processing, scoring, enrichment), write and run Python code via Bash. Don't describe what you would do — execute it and report actual results. Use ToolUniverse tools to retrieve data, then Python (pandas, scipy, statsmodels, matplotlib) to analyze it.

Domain Reasoning

PTMs are context-dependent: same phosphorylation site can activate or inhibit depending on kinase and effectors. Always check: which enzyme, what functional consequence, in what cell context.

KEY PRINCIPLES

1. Disambiguation first -- resolve to UniProt accession before iPTMnet calls
2. iPTMnet is SOAP-style -- every call requires operation parameter
3. Evidence-graded -- distinguish experimental (T1) from predicted (T4)
4. English-first queries

Workflow

Phase 0: Protein Disambiguation → UniProt accession
Phase 1: PTM Site Inventory → iPTMnet_get_ptm_sites
Phase 2: Proteoform Analysis → iPTMnet_get_proteoforms
Phase 3: PTM-Dependent Interactions → iPTMnet_get_ptm_ppi
Phase 4: Functional Context → ProtVar_get_function at key sites
Phase 4b: Linear Motif Context → ELM_get_instances for SLiM overlap
Phase 4c: Experimental Data → MassIVE/ProteomeXchange
Phase 5: Synthesis & Report

Phase 0: Disambiguation

• iPTMnetsearch(operation="search", searchterm="TP53", role="Substrate") -- find UniProt IDs
• If user provides UniProt accession directly, use it
• Select human entry if multiple hits

Phase 1: PTM Sites

iPTMnetgetptmsites(operation="getptmsites", uniprotid="P04637") -- returns position, residue, modification type, enzyme, evidence. Group by modification type. Fallback: UniProtgetentrybyaccession PTM annotations.

Phase 2: Proteoforms

iPTMnetgetproteoforms(operation="getproteoforms", uniprotid=...) -- distinct PTM combinations. Focus on those with functional/disease annotations if >20.

Phase 3: PTM-Dependent Interactions

iPTMnetgetptmppi(operation="getptmppi", uniprotid=...) -- interacting protein, PTM site, effect (enables/disrupts). Supplement with STRINGgetinteractionpartners(identifiers=gene, species=9606, requiredscore=700).

Phase 4: Functional Context

ProtVargetfunction(accession=..., position=N, variant_aa=AA) -- domain, active site, binding site, conservation. Grade: active-site PTM > domain-core > disordered region.

Phase 4b: Linear Motifs (ELM)

ELMgetinstances(operation="getinstances", uniprotid=..., motiftype="MOD") -- MOD = modification sites, DEG = degradation signals. Cross-reference with Phase 1 PTM positions. ELMlistclasses(operation="listclasses") for motif details.

Phase 4c: Experimental Data

MassIVEsearchdatasets(species="9606"), MassIVEgetdataset(accession="MSV...") for public MS datasets.

Evidence Grading

Tool Parameter Reference

Critical: All iPTMnet and ELM tools require operation as first parameter (SOAP-style).

Fallbacks

Limitations

• iPTMnet biased toward well-studied proteins
• Proteoform data covers observed combinations only
• PTM-PPI: only PTM-specific evidence; more PPIs exist in STRING