Quantitative drug-target validation pipeline. Scores druggability, selectivity, safety profile, ADMET feasibility, and structural tractability with a composite Target Validation Score (0-100) and GO/NO-GO recommendation. Use for go/no-go decisions on a target before commit-to-medchem, target priorit

Drug Target Validation Pipeline Validate drug target hypotheses using multi dimensional computational evidence before committing to wet lab work. Produces a quantitative Target Validation Score (0 100) with priority tier classification and GO/NO GO recommendation. Reasoning Before Searching A valid drug target must pass 4 gates in order. Failing an early gate makes later gates irrelevant: 1. Genetic evidence linking it to disease : Does human genetic data (GWAS, rare variant studies, Mendelian genetics) support this target's role? Genetic evidence is the strongest predictor of clinical success. Use OpenTargets and GWAS catalog before anything else. If no genetic link exists, the hypothesis is speculative — document this clearly. 2. Druggability : Can a molecule reach and modulate the target? Check structure availability (PDB, AlphaFold), binding pocket prediction (ProteinsPlus), target class (kinase, GPCR, nuclear receptor = favorable; transcription factor, scaffold protein = difficult), and existing chemical probes. 3. Safety — essentiality in normal tissue : Is the target expressed in critical tissues (heart, liver, bone marrow)? Is knockout lethal in mice? High expression in essential tissue or lethality in mouse models is a strong safety red flag even before any clinical data. 4. Competitive landscape : Are other drugs already approved or in late stage trials for this target? If so, the bar is differentiation, not first in class. Check ChEMBL, DrugBank, and ClinicalTrials.gov early. Do not proceed to Phase 3 (Chemical Matter) before completing Phase 1 (Disease Association). Gate 1 failures should prompt a NO GO or pivot recommendation. LOOK UP DON'T GUESS : Never assume a target is druggable based on its protein family alone, never assume expression is low in a tissue without checking GTEx or HPA, never assume no competitors without searching ClinicalTrials.gov. RUN THE ML MODELS, DON'T SKIP THEM : When deep learning predictors are available (ADMET AI, ESMFold, AlphaFold, DoGSite, DynaMut2, DeepGO), run them even when database lookups or experimental data already cover the same property . The ML predictions provide an orthogonal, mechanistically grounded estimate that's a first class output of this skill — not a fallback. A target validation report missing ML predictions is incomplete regardless of how much database evidence is present. 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. Key Principles 1. Report first Create report file FIRST, then populate progressively 2. Target disambiguation FIRST Resolve all identifiers before analysis 3. Evidence grading Grade all evidence as T1 (experimental) to T4 (computational) 4. Disease specific Tailor analysis to disease context when provided 5. Modality aware Consider small molecule vs biologics tractability 6. Safety first Prominently flag safety concerns early 7. Quantitative scoring Every dimension scored numerically (0 100 composite) 8. Negative results documented "No data" is data; empty sections are failures 9. Source references Every statement must cite tool/database 10. English first queries Always use English terms in tool calls; respond in user's language When to Use Apply when users ask about: "Is [target] a good drug target for [disease]?" Target validation, druggability assessment, or target prioritization Safety risks of modulating a target Chemical starting points for target validation GO/NO GO recommendation for a target Not for (use other skills): general target biology ( tooluniverse target research ), drug compound profiling ( tooluniverse drug research ), variant interpretation ( tooluniverse variant interpretation ), disease research ( tooluniverse disease research ). Input Parameters Parameter Required Description Example target Yes Gene symbol, protein name, or UniProt ID EGFR , P00533 disease No Disease/indication for context Non small cell lung cancer modality No Preferred therapeutic modality small molecule , antibody , PROTAC Reference Files SCORING CRITERIA.md Detailed scoring matrices, evidence grading, priority tiers, score calculation REPORT TEMPLATE.md Full report template, completeness checklist, section format examples TOOL REFERENCE.md Verified tool parameters, known corrections, fallback chains, modality specific guidance, phase by phase tool lists QUICK START.md Quick start guide Scoring Overview Total: 0 100 points across 5 dimensions (details in SCORING CRITERIA.md): Dimension Max Sub dimensions Disease Association 30 Genetic (10) + Literature (10) + Pathway (10) Druggability 25 Structure (10) + Chemical matter (10) + Target class (5) Safety Profile 20 Expression (5) + Genetic validation (10) + ADRs (5) Clinical Precedent 15 Based on highest clinical stage achieved Validation Evidence 10 Functional studies (5) + Disease models (5) Priority Tiers : 80 100 = Tier 1 (GO) 60 79 = Tier 2 (CONDITIONAL GO) 40 59 = Tier 3 (CAUTION) 0 39 = Tier 4 (NO GO) Evidence Grades : T1 (clinical proof) T2 (functional studies) T3 (associations) T4 (predictions) Pipeline Phases Phase 0: Target Disambiguation (ALWAYS FIRST) Resolve target to ALL identifiers before any analysis. Steps : 1. MyGene query genes Get initial IDs (Ensembl, UniProt, Entrez) 2. ensembl lookup gene Get versioned Ensembl ID (species="homo sapiens" REQUIRED) 3. ensembl get xrefs Cross references (HGNC, etc.) 4. OpenTargets get target id description by name Verify OT target 5. ChEMBL search targets Get ChEMBL target ID 6. UniProt get function by accession Function summary (returns list of strings) 7. UniProt get alternative names by accession Collision detection Output : Table of verified identifiers (Gene Symbol, Ensembl, UniProt, Entrez, ChEMBL, HGNC) plus protein function and target class. Optional fast cross database search : TargetMine search(q=<gene , facet Category="Gene") / TargetMine search genes(q=<gene ) queries one InterMine warehouse that already integrates UniProt, PDB, ChEMBL, Reactome, KEGG, DrugBank, TTD, and clinical trials data — useful as a quick single call sanity check that a gene/target resolves consistently across sources before running the full per database Phase 0 1 sequence, not a replacement for it (verified live: TargetMine search(q="EGFR", facet Category="Gene") returns real hits with organism/category facets in ~0.3s). Phase 1: Disease Association (0 30 pts) Quantify target disease association from genetic, literature, and pathway evidence. Key tools : OpenTargets get diseases phenotypes by target ensembl Disease associations OpenTargets target disease evidence Detailed evidence (needs efoId + ensemblId ) OpenTargets get evidence by datasource Evidence by data source gwas get snps for gene / gwas search studies GWAS evidence gnomad get gene constraints Genetic constraint (pLI, LOEUF) PubMed search articles Literature (returns plain list of dicts) OpenTargets get publications by target ensemblID OT publications (uses entityId ) Phase 2: Druggability (0 25 pts) Assess whether the target is amenable to therapeutic intervention. Key tools : OpenTargets get target tractability by ensemblID Tractability (SM, AB, PR, OC) OpenTargets get target classes by ensemblID Target classification Pharos get target TDL: Tclin Tchem Tbio Tdark DGIdb get gene druggability Druggability categories alphafold get prediction (param: qualifier ) / alphafold get summary ProteinsPlus predict binding sites Pocket detection OpenTargets get chemical probes by target ensemblID Chemical probes OpenTargets get target enabling packages by ensemblID TEPs TCDB get transporter For SLC/ABC transporter targets: TC classification, family, PDB structures (param: uniprot accession ) TCDB search by substrate Find transporters by substrate (param: substrate name ) Phase 2b: Kinase Specific Structural Data (KLIFS) When the target class is a protein kinase , don't stop at the generic tractability/structure checks above — KLIFS (Kinase Ligand Interaction Fingerprints and Structures) gives kinase specific binding pocket conformation data that generic PDB/AlphaFold lookups don't: KLIFS list kinases(kinase group, species) find the KLIFS kinase ID by group ( TK , TKL , STE , CK1 , AGC , CAMK , CMGC , RGC , Other ) or list all; match by gene symbol in the response KLIFS get kinase(kinase ID) kinase family/group, UniProt/IUPHAR cross refs, and the 80 residue standardized KLIFS pocket sequence KLIFS get structures(kinase ID) every annotated crystal structure for the kinase, each with DFG loop conformation ( in / out / out like ) and alphaC helix conformation ( in / out ), resolution, quality score, and bound ligand KLIFS get structures by pdb(pdb codes) same DFG/alphaC annotation, looked up by PDB code(s) instead of kinase ID (comma separated for multiple) KLIFS get ligands(kinase ID) every co crystallized inhibitor across all of that kinase's structures, with SMILES/InChIKey — the fastest way to enumerate known chemical starting points for a kinase target Why DFG/alphaC state matters for druggability scoring : DFG in + alphaC in is the active conformation (ATP competitive Type I inhibitors bind here). DFG out structures expose an allosteric back pocket adjacent to the ATP site (Type II inhibitors, e.g. imatinib) — a kinase with confirmed DFG out structures has a second , often more selective, druggable pocket beyond the ATP site, which should raise its druggability sub score. A kinase with only DFG in structures across hundreds of PDB entries may be harder to hit selectively with a Type II strategy. Report the DFG/alphaC state distribution (e.g. "527 DFG in / 23 DFG out / 15 DFG out like structures") as evidence, not just a structure count. Phase 3: Chemical Matter (feeds Phase 2 scoring) Identify existing chemical starting points for target validation. Key tools : ChEMBL search targets + ChEMBL get target activities Bioactivity data (note: target chembl id exact with double underscore) BindingDB get ligands by uniprot Binding data (affinity in nM) PubChem search assays by target gene + PubChem get assay active compounds HTS data OpenTargets get associated drugs by target ensemblID Known drugs ( size REQUIRED) ChEMBL search mechanisms Drug mechanisms DGIdb get gene info Drug gene interactions Phase 3b: ADMET AI Deep Learning Profile (REQUIRED) For each lead / approved compound identified above, run all ten ADMET AI Chemprop GNN endpoints . This is a required deliverable of the skill, not optional: Endpoint Tool Physicochemical (MW, logP, HBA/HBD, TPSA) ADMETAI predict physicochemical properties Toxicity (AMES, DILI, LD50, carcinogens, skin sensitizers, ClinTox) ADMETAI predict toxicity BBB penetrance ADMETAI predict BBB penetrance CYP interactions (1A2, 2C9, 2C19, 2D6, 3A4) ADMETAI predict CYP interactions Bioavailability (HIA, PAMPA, Caco 2, F20/F30) ADMETAI predict bioavailability Clearance & distribution (hepatocyte, microsome, VDss, PPB) ADMETAI predict clearance distribution Nuclear receptor activity (NR AR, NR AhR, NR Aromatase, NR ER, NR PPAR γ) ADMETAI predict nuclear receptor activity Stress response (SR ARE, SR ATAD5, SR HSE, SR MMP, SR p53) ADMETAI predict stress response Solubi

tooluniverse-drug-target-validation