drug-discovery
This Claude Code skill supports computational drug discovery by identifying therapeutic targets, screening compound libraries against those targets, predicting drug properties like absorption and toxicity, optimizing lead molecules through structure-activity analysis, and modeling pharmacokinetics and pharmacodynamics. Activate it when users discuss drug targets, virtual screening, ADMET prediction, lead optimization, or drug repurposing using molecular docking, cheminformatics databases, and prediction tools like SwissADME or pkCSM.
git clone --depth 1 https://github.com/beita6969/ScienceClaw /tmp/drug-discovery && cp -r /tmp/drug-discovery/skills/drug-discovery ~/.claude/skills/drug-discoverySKILL.md
## When to Trigger Activate this skill when the user mentions: - Drug target identification, druggability assessment - Virtual screening, molecular docking, pharmacophore - ADMET (absorption, distribution, metabolism, excretion, toxicity) - Lead optimization, SAR (structure-activity relationship) - Pharmacokinetics (PK), pharmacodynamics (PD), PK/PD modeling - Drug repurposing, off-label, drug-disease associations - SMILES, InChI, compound libraries, chemical fingerprints - IC50, EC50, Ki, dose-response curves ## Step-by-Step Methodology 1. **Target identification and validation** - Identify therapeutic target from literature, GWAS hits, or omics data. Assess druggability using Open Targets, DGIdb, or structural pocket analysis. Confirm target-disease association strength. 2. **Compound sourcing** - Search ChEMBL, PubChem, ZINC, or DrugBank for known active compounds. For novel scaffolds, consider de novo design tools (REINVENT, MolGPT). 3. **Virtual screening** - Structure-based: dock compound library against target (AutoDock Vina, Glide). Ligand-based: use pharmacophore models or molecular fingerprint similarity. Filter by drug-likeness (Lipinski Ro5, Veber rules). 4. **ADMET prediction** - Predict absorption (Caco-2 permeability, logP), distribution (plasma protein binding, Vd), metabolism (CYP inhibition/induction), excretion (clearance), and toxicity (hERG, hepatotoxicity, AMES mutagenicity). Use SwissADME, pkCSM, or ADMETlab. 5. **Lead optimization** - Analyze SAR from dose-response data. Identify key pharmacophoric features. Suggest modifications to improve potency, selectivity, or ADMET profile while maintaining drug-likeness. 6. **PK/PD modeling** - Build compartmental PK models. Estimate key parameters: Cmax, Tmax, AUC, half-life, bioavailability. For PD, model dose-response (Emax model, Hill equation). 7. **Drug repurposing analysis** - Query drug-gene interaction databases. Analyze shared pathways between drug targets and disease mechanisms. Check clinical trial databases for existing evidence. ## Key Databases and Tools - **ChEMBL** - Bioactivity data for drug-like compounds - **PubChem** - Chemical structure and bioassay data - **DrugBank** - Drug and target information - **Open Targets** - Target-disease associations - **ZINC** - Purchasable compound library - **SwissADME / pkCSM** - ADMET prediction tools - **BindingDB** - Protein-ligand binding data ## Output Format - Compound results as tables: SMILES, molecular weight, logP, key activity (IC50/EC50), ADMET flags. - Docking results: binding energy (kcal/mol), key interactions, pose description. - PK parameters: Cmax, Tmax, AUC, t1/2, clearance, bioavailability with units. - SAR analysis: matched molecular pair comparisons with activity changes. ## Quality Checklist - [ ] Target-disease association supported by evidence (genetic, functional) - [ ] Drug-likeness filters applied (Lipinski, Veber, PAINS) - [ ] ADMET predictions include confidence levels or applicability domain - [ ] Docking validated against known co-crystal structures when available - [ ] IC50/EC50 reported with assay conditions and confidence intervals - [ ] PK parameters include units and species (human vs. preclinical) - [ ] Known liabilities (hERG, CYP inhibition, reactive metabolites) flagged - [ ] Comparison to existing drugs/compounds for the same target included
Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/running secrets via op.
Transparent, rigorous research with full methodology — not a black-box API wrapper. Conducts exhaustive investigation through mandated 2-cycle research per theme, APA 7th citations, evidence hierarchy, and 3 user checkpoints. Self-contained using native OpenClaw tools (web_search, web_fetch, sessions_spawn). Use for literature reviews, competitive intelligence, or any research requiring academic rigor and reproducibility.
# Academic Literature Search — 学术文献检索与引用管理
Route plain-language requests for Pi, Claude Code, Codex, OpenCode, Gemini CLI, or ACP harness work into either OpenClaw ACP runtime sessions or direct acpx-driven sessions ("telephone game" flow). For coding-agent thread requests, read this skill first, then use only `sessions_spawn` for thread creation.
Manage Apple Notes via the `memo` CLI on macOS (create, view, edit, delete, search, move, and export notes). Use when a user asks OpenClaw to add a note, list notes, search notes, or manage note folders.
Manage Apple Reminders via remindctl CLI (list, add, edit, complete, delete). Supports lists, date filters, and JSON/plain output.
Write articles, guides, blog posts, tutorials, newsletter issues, and other long-form content in a distinctive voice derived from supplied examples or brand guidance. Use when the user wants polished written content longer than a paragraph, especially when voice consistency, structure, and credibility matter.
Search and retrieve preprints from arXiv via the Atom API. Use this skill when searching for papers in physics, mathematics, computer science, quantitative biology, quantitative finance, statistics, electrical engineering, or economics by keywords, authors, arXiv IDs, date ranges, or categories.