msa-structure-prediction-pipeline

Predicting protein structure using MSA search results involves chaining MSA-Search and OpenFold3 NIM microservices to retrieve evolutionary homologs from UniRef30 and generate high-confidence PDB structural models.

MSA-informed protein structure prediction is a process that uses multiple sequence alignment data from evolutionary homologs to guide folding algorithms, ensuring high-accuracy structural models for single proteins and multi-chain complexes.

Yes, you need valid NGC API credentials and access to BioNeMo NIM endpoints to run the protein structure prediction pipeline, unless you configure the skill to use local Docker deployments for data privacy.

Yes, you can predict multi-chain protein complexes with OpenFold3 through this pipeline, which facilitates the generation of MSA-informed structural models for both single proteins and multi-chain complexes.

Yes, you can use local Docker deployments for protein folding to ensure data privacy, as the skill supports flexible configuration for both hosted NVIDIA NIM endpoints and local Docker environments.