AIPS BOOM

NIM-for-AlphaFold2

2 min read

NIM for AlphaFold2

Type: Microservice Tags: NVIDIA, NIM, BioNeMo, AlphaFold2, protein structure prediction, MSA, drug discovery, computational biology Related: NVIDIA-BioNeMo, NVIDIA-NIM, NIM-for-MSA-Search, NIM-for-AlphaFold2-Multimer, NIM-for-OpenFold2, NIM-for-OpenFold3, NIM-for-Boltz2, NIM-for-ProteinMPNN, NIM-for-Evo-2, NVIDIA-Clara, NVIDIA-AI-Enterprise, NGC Sources: https://docs.nvidia.com/nim/bionemo/alphafold2/latest/overview.html; https://docs.nvidia.com/nim/bionemo/alphafold2/latest/index.html Last Updated: 2026-04-29

Summary

NIM for AlphaFold2 is NVIDIA’s NIM microservice for AlphaFold2 protein structure prediction. It predicts protein structure from an amino-acid sequence, can generate a multiple sequence alignment (MSA), and can predict structure from a pre-computed MSA. In the wiki, this is the single-chain AlphaFold2 BioNeMo NIM page.

Detail

The current NVIDIA docs describe AlphaFold2 as a Google DeepMind protein structure prediction model with state-of-the-art CASP performance. NVIDIA’s AlphaFold2 NIM packages this capability behind a consistent NIM API and includes GPU-accelerated MMSeqs2 for multiple sequence alignment.

This page should be used when a question asks for NVIDIA’s AlphaFold2 deployment, API, quickstart, performance, or runtime configuration. Use NIM-for-AlphaFold2-Multimer for multiple interacting protein sequences, NIM-for-OpenFold2 for the OpenFold PyTorch reimplementation, and NIM-for-OpenFold3 or NIM-for-Boltz2 for more general biomolecular complex prediction.

Connections

  • NVIDIA-BioNeMo - BioNeMo is the NVIDIA life-sciences platform that includes AlphaFold2 NIM.
  • NVIDIA-NIM - AlphaFold2 is delivered as a NIM microservice.
  • NIM-for-MSA-Search - GPU-accelerated MSA and template search can feed AlphaFold2-style structure prediction workflows.
  • NIM-for-AlphaFold2-Multimer - multimer variant for multiple protein sequences and complexes.
  • NIM-for-OpenFold2 - OpenFold2 is an open PyTorch reimplementation of AlphaFold2 concepts.
  • NIM-for-OpenFold3 - extends structure prediction toward all-atom biomolecular complexes.
  • NIM-for-Boltz2 - newer biomolecular structure and binding-affinity NIM.
  • NIM-for-Evo-2 - DNA foundation model NIM in the BioNeMo ecosystem.
  • NIM-for-ProteinMPNN - protein sequence design NIM adjacent to structure prediction/validation workflows.
  • NVIDIA-Clara - Clara is the broader healthcare/life-sciences platform.
  • NVIDIA-AI-Enterprise - enterprise support and security context for production NIM deployments.
  • NGC - distribution path for NIM containers and model assets.

Source Excerpts

  • “Predict protein structure”
  • “GPU-accelerated MMSeqs2”
  • “consistent API”

Graph View

Backlinks