Skip to content
FyreFlySystems

Structure and complexes

Predict how a chain folds, alone or together with a ligand.

Provenance reads built by · served by. Models served through NVIDIA BioNeMo NIM process your input on NVIDIA infrastructure, which may be in the United States — see the privacy policy. NVIDIA’s hosted models are used only in the Research Preview, for testing; paid plans run on machines operated by FyreFly Systems.

Abstract 3D render of a cluster of translucent glass spheres in blue, violet and coral, linked by thin stems Google DeepMind / Pexels ↗

Models in this category

Boltz-2

Predicts the 3D structure of a protein, DNA or RNA chain, alone or together with a small-molecule ligand, and can estimate the ligand’s binding affinity.

Takes
  • Protein, DNA or RNA sequence
  • Ligand as SMILES or CCD code, optional
Returns
  • Structure (mmCIF)
  • pLDDT and pTM
  • Binding affinity, optional
Per job
1–5 structures · 1–6 recycling steps · 10–1,000 sampling steps · affinity for one ligand
Confidence
pLDDT 0–100 per residue · pTM 0–1
Built · served by
MIT Jameel Clinic · NVIDIA BioNeMo NIM
Reference
Repository ↗

OpenFold3

Predicts biomolecular structures of protein, DNA or RNA chains, optionally with a ligand, with or without a multiple sequence alignment.

Takes
  • Protein, DNA or RNA sequence
  • MSA as A3M or CSV, optional
  • Ligand as SMILES or CCD code, optional
Returns
  • Structure (PDB or mmCIF)
  • pTM and ipTM
Per job
1–5 structures
Confidence
pTM 0–1 · ipTM 0–1 for the interfaces between chains
Built · served by
OpenFold Consortium · NVIDIA BioNeMo NIM