Design antigen-conditioned antibody and nanobody CDR loops from X-masked FASTA inputs.
npx skills add https://github.com/junior1p/ProteinClaw --skill iggm-junior1p
Please help me install this Agent Skill. Skill: iggm Source: https://github.com/junior1p/ProteinClaw/tree/main/skills/iggm Command: npx skills add https://github.com/junior1p/ProteinClaw --skill iggm-junior1p
Antibody and nanobody CDR design is time-consuming and requires specialized models to produce antigen-conditioned, diverse, and structurally compatible loop sequences. IgGM streamlines CDR generation and redesign by taking X-masked FASTA inputs and a target antigen structure to produce candidate sequences and co-folded complexes ready for downstream validation.
Design 20 nanobody CDR variants against antigen.pdb chain A using an X-marked FASTA and request FASTA, PDB, and scores for each design.
High-intent search queries and answers about installing and using this skill.
To generate antigen-conditioned antibody CDRs, you provide an X-masked FASTA sequence and a target antigen PDB chain, and the system outputs designed FASTA sequences, co-folded complex PDBs, and confidence scores.
Yes, you can design de novo VHH nanobody CDR loops by providing an X-marked FASTA scaffold alongside the target antigen PDB chain for epitope-targeted generation.
Yes, you need GPU resources to run antibody CDR design and co-folding tasks, with a recommended 24GB VRAM for model execution and complex generation.
The best way to redesign specific CDR loops while preserving framework residues is to mask only the target CDR positions with X in the FASTA input and condition generation on the target antigen PDB.
Yes, antibody CDR generation supports joint heavy and light chain CDR design conditioned on an antigen epitope for structure-informed binder discovery.
You validate generated antibody complex PDBs by passing the output FASTA and PDB files to downstream validation skills like chai and protenix, and applying quality filters like protein-qc.
ovachieverGenerate protein sequences, structures, and embeddings using ESM3 and ESM C models.
OwnLabAIGenerate protein-ligand docking poses with diffusion models and confidence scores.
adaptyvbioPredict protein structures from FASTA sequences using Chai-1.
junior1pDesign VHH nanobody binders against target proteins using structure-conditioned generation.
