AI-driven patient-to-trial matching for precision medicine and oncology. Given a patient profile (disease, molecular alterations, stage, prior treatments), discovers and ranks clinical trials from ClinicalTrials.gov using multi-dimensional matching across molecular eligibility...
78%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
Fix and improve this skill with Tessl
tessl review fix ./scientific-skills/Evidence Insight/tooluniverse-clinical-trial-matching/SKILL.mdTransform patient molecular profiles and clinical characteristics into prioritized clinical trial recommendations. Searches ClinicalTrials.gov and cross-references with molecular databases (CIViC, OpenTargets, ChEMBL, FDA) to produce evidence-graded, scored trial matches.
KEY PRINCIPLES:
Apply when user asks:
NOT for (use other skills instead):
tooluniverse-cancer-variant-interpretationtooluniverse-adverse-event-detectiontooluniverse-drug-target-validationtooluniverse-disease-research| Input Format | Parsed As | Example |
|---|---|---|
| Gene + amino acid change | Specific mutation | EGFR L858R |
| Gene + exon notation | Exon-level alteration | EGFR exon 19 deletion |
| Gene + fusion partner | Fusion | EML4-ALK fusion |
| Gene + amplification | Copy number gain | HER2 amplification |
| Gene + expression level | Expression biomarker | PD-L1 50% |
| Gene + status | Status biomarker | MSI-high, TMB-high |
| Gene + resistance | Resistance mutation | EGFR T790M |
| Common Alias | Official Symbol | Notes |
|---|---|---|
| HER2 | ERBB2 | Search both in trials |
| PD-L1 | CD274 | Often searched as "PD-L1" in trials |
| ALK | ALK | EML4-ALK is a fusion |
| VEGF | VEGFA | Often searched as "VEGF" |
| PD-1 | PDCD1 | Search as "PD-1" in trials |
| BRCA | BRCA1/BRCA2 | Specify which BRCA gene |
Detailed parsing rules and regex patterns: see references/parsing_and_validation.md
| Phase | Name | Summary |
|---|---|---|
| 0 | Tool Parameter Reference | Verify all tool parameters before calling. See references/phases_detail.md |
| 1 | Patient Profile Standardization | Resolve disease→EFO ID, genes→Ensembl/Entrez IDs, classify biomarker actionability |
| 2 | Broad Trial Discovery | Disease/biomarker/intervention searches on ClinicalTrials.gov, deduplicate results |
| 3 | Trial Characterization | Batch-fetch eligibility, interventions, locations, status, descriptions for candidate NCT IDs |
| 4 | Molecular Eligibility Matching | Parse eligibility text for biomarker requirements, score patient-trial molecular match (0-40) |
| 5 | Drug-Biomarker Alignment | Identify trial drug mechanisms via OpenTargets/ChEMBL, verify target overlap with patient biomarkers |
| 6 | Evidence Assessment | FDA approval, PubMed literature, CIViC evidence, evidence tier classification (T1-T4) |
| 7 | Geographic & Feasibility | Trial site locations, enrollment status, proximity to patient location |
| 8 | Alternative Options | Basket/tumor-agnostic trials, expanded access, compassionate use programs |
| 9 | Scoring & Ranking | Calculate Trial Match Score (0-100), assign tier, rank trials |
| 10 | Report Synthesis | Generate markdown report with executive summary, ranked trials, evidence grading, checklist |
Detailed phase execution code and tool call examples: see references/phases_detail.md
| Component | Max Points | Key Criteria |
|---|---|---|
| Molecular Match | 40 | Exact variant match=40, Gene-level=30, Pathway=20, No criteria=10, Excluded=0 |
| Clinical Eligibility | 25 | All criteria met=25, Most=18, Some=10, Ineligible=0 |
| Evidence Strength | 20 | FDA-approved (T1)=20, Phase III (T2)=15, Phase II (T3)=10, Phase I (T4)=5 |
| Trial Phase | 10 | Phase III=10, Phase II=8, Phase I/II=6, Phase I=4 |
| Geographic Feasibility | 5 | Patient's city=5, Same country=3, International=1, Unknown=0 |
| Score | Tier | Label | Action |
|---|---|---|---|
| 80-100 | Tier 1 | Optimal Match | Strongly recommend - contact site immediately |
| 60-79 | Tier 2 | Good Match | Recommend - discuss with care team |
| 40-59 | Tier 3 | Possible Match | Consider - needs further eligibility review |
| 0-39 | Tier 4 | Exploratory | Backup option - consider if Tier 1-3 unavailable |
Detailed scoring rules, evidence tier definitions, and matching algorithms: see references/scoring_and_matching.md
Report file: clinical_trial_matching_[DISEASE]_[BIOMARKER]_[DATE].md
Full report template with markdown structure: see references/phases_detail.md#phase-10-report-synthesis
civic_search_variants/civic_search_evidence_items do NOT filter by query. Use civic_get_variants_by_gene with gene ID instead.Full edge case handling and use patterns: see references/parsing_and_validation.md
| Gene | CIViC ID | Gene | CIViC ID |
|---|---|---|---|
| ALK | 1 | MET | 52 |
| ABL1 | 4 | PIK3CA | 37 |
| BRAF | 5 | ROS1 | 118 |
| EGFR | 19 | RET | 122 |
| ERBB2 | 20 | NTRK1 | 197 |
| KRAS | 30 | NTRK2 | 560 |
| TP53 | 45 | NTRK3 | 561 |
This skill accepts requests that match the documented purpose of tooluniverse-clinical-trial-matching and include enough context to complete the workflow safely.
Do not continue the workflow when the request is out of scope, missing a critical input, or would require unsupported assumptions. Instead respond:
tooluniverse-clinical-trial-matchingonly handles its documented workflow. Please provide the missing required inputs or switch to a more suitable skill.
| File | Content |
|---|---|
| references/phases_detail.md | Phase 0-10 detailed execution code, tool call examples, parameter tables, report template |
| references/scoring_and_matching.md | Scoring algorithm details, drug-biomarker alignment rules, evidence tier classification, matching logic |
| references/parsing_and_validation.md | Biomarker parsing regex, eligibility text parsing, gene resolution code, edge case handling, common use patterns |
aipoch/medical-research-skills
686e09d
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.