Generate protein backbones and scaffolds via diffusion models.
npx skills add https://github.com/yjcyxky/biominer-skills --skill rfdiffusion-yjcyxky
Please help me install this Agent Skill. Skill: rfdiffusion Source: https://github.com/yjcyxky/biominer-skills/tree/main/rfdiffusion Command: npx skills add https://github.com/yjcyxky/biominer-skills --skill rfdiffusion-yjcyxky
RFdiffusion provides a diffusion-model-based approach to generating novel protein backbones and designs, addressing the challenge of de novo design, motif scaffolding, binder design, and symmetric assembly in protein engineering.
Install Docker and run RFdiffusion with a default 100-residue test design: docker run --rm --gpus all -v /data/rfdiffusion:/data ghcr.io/open-prophetdb/rfdiffusion:arm64-blackwell inference.output_prefix=/data/outputs/test inference.model_directory_path=/data/models 'contigmap.contigs=[100-100]' inference.num_designs=10
High-intent search queries and answers about installing and using this skill.
To generate protein backbones for de novo design, RFdiffusion applies a diffusion model to create novel protein scaffolds. It executes GPU-accelerated runs via Docker to output structural backbones for computational biology workflows.
Motif scaffolding and binder design require a GPU-enabled Docker environment and pre-downloaded model checkpoints. You must mount local data volumes to the Docker container to execute RFdiffusion and store generated protein structures.
Yes, you can design symmetric oligomers using diffusion-based protein generation. RFdiffusion explicitly supports symmetric assembly design, enabling the creation of oligomeric protein structures for protein engineering and therapeutics.
RFdiffusion integrates with ProteinMPNN for coupled sequence design and validation workflows. This combination allows you to validate generated protein backbones by designing and testing their corresponding amino acid sequences.
The best way to start a 100-residue test design is running the provided RFdiffusion Docker command with the 'contigmap.contigs=[100-100]' parameter. This GPU-accelerated Docker execution generates ten default design outputs in your mounted data volume.
YuzhengShiDiagnose containerized agent execution issues by analyzing logs and verifying configurations.
PKU-YuanGroupConverts PDF documents to clean, structured Markdown.
adaptyvbioDesign protein sequences from structural backbones using ProteinMPNN inverse folding.
zongtingweiGenerate de novo protein backbones with a diffusion-based generative model.
