AA
AAV vector design: capsid selection, promoter optimization, payload capacity.
FreedomIntelligence/OpenClaw-Medical-Skills查看
数据与复现
MO
molecule-evolution-agent
维护者 FreedomIntelligence · 最近更新 2026年4月1日
Directed molecular evolution: generative models for compound optimization and library design.
原始来源
FreedomIntelligence/OpenClaw-Medical-Skills
https://github.com/FreedomIntelligence/OpenClaw-Medical-Skills/tree/main/skills/molecule-evolution-agent
- 维护者
- FreedomIntelligence
- 许可
- MIT
- 最近更新
- 2026年4月1日
技能摘要
来自 SKILL.md 的关键信息
核心说明
- Molecule Evolution Agent acts as autonomous medicinal chemist. It takes starting molecule (或 uses default like Aspirin) 、 iteratively modifies its structure to optimize binding ，用于 specific protein target。
- Lead Optimization：When you have hit molecule 、 want to improve its potency。
- De Novo Design：To explore chemical space around target protein。
- Idea Generation：To get creative structural modifications suggested by LLM。
- Lead Optimization：When you have hit molecule 、 want to improve its potency. De Novo Design：To explore chemical space around target protein. * Idea Generation：To get creative structural modifications suggested by LLM。
原始文档
SKILL.md 摘录
Core Capabilities
- SMILES Manipulation: Reads and writes chemical structures in SMILES format.
- LLM Chemist: Uses an LLM to suggest chemically valid modifications (e.g., "Add a fluorine group to the ring").
- Mock Scoring: (Currently) Uses a mock scoring function to simulate docking affinity.
Workflow
- Input: Target Protein Name (e.g., "GPRC5D").
- Process:
- Start with a seed molecule.
- Loop for N generations.
- Ask LLM for a modification.
- Score the new molecule.
- Keep the best candidate.
- Output: Top candidate SMILES and the evolution history.
Example Usage
User: "Design a better binder for GPRC5D."
Agent Action:
python3 Skills/Drug_Discovery/Molecule_Design/evolution_agent.py
适用场景
- **Lead Optimi。
不适用场景
- Do not rely on this catalog entry alone ，用于 installation 或 maintenance details。
相关技能
相关技能
AG
AgentD autonomous drug discovery: target identification, hit finding, ADMET optimization.
FreedomIntelligence/OpenClaw-Medical-Skills查看
Call HLA alleles from NGS data using OptiType, HLA-HD, or arcasHLA for immunogenomics applications. Use when determining…
FreedomIntelligence/OpenClaw-Medical-Skills查看
Query PharmGKB and CPIC for drug-gene interactions, pharmacogenomic annotations, and dosing guidelines. Use when predict…
FreedomIntelligence/OpenClaw-Medical-Skills查看