Automate end-to-end de novo protein design workflows with YAML experiment schemas.
npx skills add https://github.com/LiorZ/protein-design-skills --skill genie3
Please help me install this Agent Skill. Skill: genie3 Source: https://github.com/LiorZ/protein-design-skills/tree/main/skills/genie3 Command: npx skills add https://github.com/LiorZ/protein-design-skills --skill genie3
Genie3 automates end-to-end de novo protein design workflows from installation to design evaluation, enabling researchers to explore scaffolded designs and binders at scale.
Clone the Genie 3 repo, install the environment, and run a tiny unconditional example with the included YAML.
High-intent search queries and answers about installing and using this skill.
End-to-end de novo protein design automates workflows from installation to evaluation, supporting unconditional monomer design, motif scaffolding, and binder design with multi-GPU scaling and CLI-based experiment management.
SE(3)-equivariant diffusion generates protein structures by modeling geometric transformations that preserve spatial equivariance, enabling fast end-to-end design of monomers, motif scaffolds, and target binders.
Yes, binder design against targets integrates with backends like ProteinMPNN and ColabFold, supporting beam search, iterative conditioning, and result-reduction tooling to produce publishable designs.
Motif scaffolding supports multi-GPU and multi-node scaling through a CLI interface that manages experiment execution, generation, evaluation, and training using enforced YAML-based experiment schemas.
Yes, the system enforces a YAML-based experiment schema for all workflows including unconditional monomer design, providing a quick start path with included YAML examples and ready-made problem sets.
Diffusion-based protein design workflows require multi-GPU infrastructure for scaling and rely on YAML schema enforcement, with results depending on backend integrations like ProteinMPNN for sequence design and ColabFold for validation.
adaptyvbioGenerate protein backbones with RFdiffusion using contigmap and hotspot specifications.
zongtingweiOrchestrate end-to-end binder design workflows with AF2 validation and QC metrics.
PKU-YuanGroupConverts PDF documents to clean, structured Markdown.
heimeshoffOrchestrate end-to-end F# feature development across shared contracts, backend, frontend, and tests.
