Predict 3D structures for protein, DNA, RNA, ligand, and ion complexes.
npx skills add https://github.com/junior1p/ProteinClaw --skill protenix
Please help me install this Agent Skill. Skill: protenix Source: https://github.com/junior1p/ProteinClaw/tree/main/skills/protenix Command: npx skills add https://github.com/junior1p/ProteinClaw --skill protenix
Predict high-accuracy 3D structures for protein, nucleic acid, ligand, and ion complexes to validate designs and inform experimental prioritization, while providing confidence metrics for ranking candidates.
Run Protenix on my complex.faa with seeds 42,43,44 using no-use-msa to generate CIF models and a confidence JSON for ranking.
High-intent search queries and answers about installing and using this skill.
You can predict 3D structures for protein, DNA, RNA, ligand, and ion complexes by submitting FASTA or Protenix JSON inputs to generate CIF structure files and confidence JSON metrics for validation and experimental prioritization.
MSA-free mode allows you to run rapid predictions without multiple sequence alignments, enabling high-throughput structural campaigns while still producing CIF models and confidence metrics for evaluating complex designs.
To generate multi-seed ensembles for robust confidence estimates, you specify multiple seeds during your structure prediction run, which produces varied CIF models and confidence metrics to help validate your protein complex designs.
Yes, protein-ligand complex structure prediction supports SMILES inputs for ligand entities alongside protein, DNA, RNA, and ion entities, outputting CIF structure files and confidence JSON metrics for downstream analysis.
AF3-quality structure prediction via this PyTorch implementation recommends GPU acceleration with 24 to 48GB VRAM to efficiently process single-chain and multi-chain complex predictions and generate CIF output files.
Structure prediction outputs a confidence JSON file containing pLDDT, pTM, and ipTM metrics, which are used for ranking candidates, filtering designs, and integrating with downstream QC and campaign skills.
001TMFPredict protein, complex, and protein-ligand structures from sequence inputs with confidence metrics.
zongtingweiFilter protein-design candidates using structural, interface, and biophysical QC metrics.
NVIDIA-BioNeMoChain MSA-Search and OpenFold3 NIM microservices to predict protein structures.
pradyumnasagarPredicts protein structures with AlphaFold 2/3 and Boltz-1 from sequence input.
