ESMFold
MSA-free structure prediction from a protein language model
ESMFold predicts 3D protein structures directly from single sequences using a large protein language model, eliminating MSA database searches. Predictions run orders of magnitude faster than MSA-based pipelines while retaining useful accuracy on single-domain proteins. Ideal for rapid folding of metagenomic proteins and large sequence libraries.
ESMFold is being onboarded — request access and be first in line.
At a glance
- Input
- Protein sequence (FASTA)
- Output
- Predicted structure (PDB)
- Developed by
- Meta AI
- Published
- Lin et al., Science 2023 · 2022
#msa-free#language-model#fast
Related tools
More methods in Structure Prediction & Folding
