proteinmpnn
关于
The `proteinmpnn` skill performs inverse folding to design protein sequences for given backbones, ideal for redesigning sequences or optimizing for stability. Key features include fixing specific residues during design and supporting multi-state/negative design scenarios. Use `rfdiffusion` for backbone generation and `ligandmpnn`/`solublempnn` for specialized ligand-aware or solubility tasks.
快速安装
Claude Code
推荐npx skills add NeverSight/skills_feed -a claude-code/plugin add https://github.com/NeverSight/skills_feedgit clone https://github.com/NeverSight/skills_feed.git ~/.claude/skills/proteinmpnn在 Claude Code 中复制并粘贴此命令以安装该技能
GitHub 仓库
Frequently asked questions
What is the proteinmpnn skill?
proteinmpnn is a Claude Skill by NeverSight. Skills package instructions and resources that Claude loads on demand, so Claude can perform proteinmpnn-related tasks without extra prompting.
How do I install proteinmpnn?
Use the install commands on this page: add proteinmpnn to Claude Code as a plugin, or clone its repository into your skills directory, then restart Claude so it picks up the skill.
What category does proteinmpnn belong to?
proteinmpnn is in the design-tools category, tagged sequence-design and inverse-folding.
Is proteinmpnn free to use?
Yes. proteinmpnn is listed on AIMCP and free to install. It runs inside Claude, so no separate service account is required to use the skill itself.
相关推荐技能
Boltz是一个开源的生物分子结构预测工具，用于预测蛋白质复合物和蛋白质-配体复合物的三维结构。它特别适合作为AlphaFold2的开源替代方案，可用于验证设计的结合剂，并能利用本地GPU资源进行计算。开发者可在需要结构预测但希望使用开源工具时调用此技能。
AlphaFold技能使用AlphaFold2进行蛋白质结构预测验证，适用于设计序列折叠验证、结合物-靶标复合物结构预测以及多链复合物预测等场景。它能计算pLDDT、pTM和ipTM等关键置信度指标，支持设计自洽性验证。对于更快的单链预测建议使用esm技能，质量控制阈值可参考protein-qc技能。
BoltzGen是一个用于全原子蛋白质设计的扩散模型技能，特别适用于需要从一开始就考虑侧链的设计任务。它在设计结合小分子或配体的蛋白质、要求精确结合几何构型时表现出色，支持基于YAML的配置。当需要纯骨架生成或序列设计时，建议分别使用rfdiffusion和proteinmpnn技能。
BindCraft是一个端到端的蛋白质结合剂设计工具，通过联合骨架和序列优化生成结合剂。它内置了AF2验证功能，提供快、中、慢三种设计协议，适用于生产级设计任务并追求高实验成功率。当需要完整设计流程时使用此工具，而仅需骨架生成时可选用rfdiffusion。