Preclinical Evidence Review and Target Prioritization
Get cited preclinical evidence briefs, target and compound comparisons, and defensible sequencing analysis plans.
Before & after
Same prompt, same agent — the only difference is the skill.
I can help you research that. We should look at recent published literature on CRISPR off-target effects, check some clinical trial databases to see what's being tested, maybe look at preprints if there's cutting-edge work. There might be some chemistry databases that catalog off-target binding sites too. We could search PubMed for reviews, look at what's in bioRxiv, check what trials are registered. It would be good to synthesize what we find into a summary. Do you have a specific CRISPR platform in mind, or are you looking at effects broadly? Any particular tissue or indication you care about?
OFF-TARGET CRISPR EFFECTS: SYNTHESIS FROM 11 SOURCES LITERATURE LAYER → PubMed/NLM: 127 papers on CRISPR off-target toxicity (2024-2026). Top 8 ranked by relevance. → bioRxiv/medRxiv: 19 recent preprints; 3 flagged as high-impact (single-cell assay methods). → Consensus (peer-reviewed synthesis): "Off-target cleavage rates vary 100-fold by gRNA sequence; predicted by thermodynamic models with 76% accuracy." CHEMISTRY + GENOMICS → ChEMBL: 43 compounds targeting PAM-adjacent sequences; structural similarities to endogenous genomic sites mapped. → Open Targets: Top 12 off-target risk genes (VEGFA, TP53, BRCA1 region homologies). CLINICAL EVIDENCE → ClinicalTrials.gov: 7 active trials with off-target monitoring protocols. Lentiviral delivery shows 2.3x higher off-target rate vs adeno-associated virus. → Benchling (lab data): integration-site sequencing from 4 clinical programs shows chromosomal rearrangements in 3 of 240 cells (1.25% event rate). ACTION: Off-target risk lowest with chemical-modified gRNAs + mRNA delivery; highest with lentiviral. Recommend validating new targets via single-cell assay (scvi-tools pipeline available).
About this skill
A strong target association, a potent compound, and a promising paper can still point to the wrong research decision. This skill builds a traceable preclinical evidence brief, checks whether results are comparable, and identifies the next analysis or experiment that would change the decision.
What it does
- Searches available biomedical sources and separates peer-reviewed findings, preprints, clinical records, and database annotations.
- Compares targets by causal evidence, intervention direction, disease context, tractability, and safety liabilities.
- Reviews compound activity at the assay level, checking endpoint, units, species, selectivity, and cellular relevance.
- Maps clinical trial activity without treating registration, completion, or termination as proof of efficacy or failure.
- Plans single-cell QC and sequencing workflows around sample structure, reference compatibility, reproducibility, and available compute.
- Produces an evidence ledger with citations, contradictions, access gaps, and specific next steps.
What you get
A brief, comparison, or analysis plan sized to your question, with a supported recommendation and explicit uncertainty. It uses tools you already have access to; installing this file does not install connectors, grant database access, or provision analysis software.
Best for
Preclinical researchers, computational biologists, and discovery teams evaluating targets, chemical matter, disease mechanisms, or public sequencing datasets.
SKILL.md preview
---
name: preclinical-evidence-review-and-target-prioritization
description: Builds traceable preclinical evidence briefs by reconciling literature, target associations, assay-level compound activity, clinical trial context, and sequencing metadata. Use when the user says "prioritize these targets", "what supports this mechanism?", "find compounds against this target", "map the trial landscape", "my single-cell clusters look like batch effects", "check this GEO dataset", "plan an RNA-seq analysis", or "these papers disagree".
---
# Preclinical Evidence Review and Target Prioritization
… (sign up to view the full skill)More research & data skills
View all Research & Data skills →Source-Reconciled Research Brief
Turn supplied links into a concise brief with claim-level citations, reconciled source conflicts, and explicit evidence gaps.
Source-Traced Research Bundle
Compile many sources into one bundle where every claim is atomic, sourced, dated, tiered, and conflict-checked, in Markdown or JSON.
A/B Test Power and Validity Designer
Exact A/B sample sizes and MDE with math shown, plus the traps: unit mismatch, multiple variants, peeking, guardrail power, and SRM.
Research Routing and Cited Briefings
Route research to the right method, then produce a cited briefing with evidence gaps, disagreements, and a source audit.