AlphaFold2
Atomic-accuracy protein and complex structures from sequence
Takes amino acid sequences, optionally with multiple sequence alignments, and predicts 3D structures with near-experimental accuracy using an attention-based neural network with Evoformer and structure modules. Handles monomers and multimeric complexes. Outputs a PDB model with per-residue pLDDT confidence and PAE maps.
AlphaFold2 is live on FoldForge — paste a sequence and press Fold.
At a glance
- Input
- Protein sequence (FASTA)
- Output
- Predicted structure (PDB)
- Developed by
- Google DeepMind
- Published
- Jumper et al., Nature 2021 · 2021
#deep-learning#msa#complexes
Related tools
More methods in Structure Prediction & Folding
