π Table of contents
π‘ Notes
- The following lists are curated by humans, as such may be incomplete
- We only include software targeting the folding problem combining learnings from AlphaFold 2 and protein language models. You may find other ML on protein tools at Kevin's incredible ML for proteins list.
- We do not wish to advertize one tool over any other, but simply list the tools we are aware of in either random or alphabetical order
- Any suggestions for improvements and additions are welcome as issues or pull requests
- Projects we identify as discontinued are marked with π
β‘οΈ Brought to you by:
[in alphabetical order]
-
MSA-based (uses Multiple Sequence Alignment as input)
- AlphaFold2 (JAX) - OG
- Manuscript: https://www.nature.com/articles/s41586-021-03819-2
- Code: https://github.com/deepmind/alphafold
- Other: Colab Notebook
- ColabFold (JAX) - faster AF2 compiling and MSA generations
- FastFold (Pytorch) - runtime improvements to OpenFold (see below)
- Manuscript: https://arxiv.org/abs/2203.00854
- Code: https://github.com/hpcaitech/FastFold
- HelixFold (PaddlePaddle) - reimplements AF2 in PaddlePaddle
- MineSpore-Fold (mindspore) - reimplements AF2 in MindSpore
- OpenFold (PyTorch) - reimplements AF2 in PyTorch; provides training code and new model params
- Manuscript: N/A
- Code: https://github.com/aqlaboratory/openfold
- Other: Colab Notebook
- RoseTTAFold (PyTorch) - reproduction attempt of AF2 in PyTorch, before details of AF2 were available; new model params
- Uni-Fold (PyTorch/JAX) - reimplements AF2 in PyTorch; provides training code and new (monomer/multimer) model params
- Manuscript: https://doi.org/10.1101/2022.08.04.502811
- Code (PyTorch): https://github.com/dptech-corp/Uni-Fold
- Code (JAX): https://github.com/dptech-corp/Uni-Fold-jax
- π Lucidrains' AlphaFold2 (PyTorch) - AF2 reproduction attempt
- π Lupoglaz's OpenFold2 (PyTorch) - AF2 reproduction attempt
- AlphaFold2 (JAX) - OG
-
pLM-based (uses Protein Language Model as input)
- ESM-Fold (PyTorch)
- Manuscript: https://doi.org/10.1101/2022.07.20.500902
- Code: N/A
- Other: [tweet] Alex's announcement,
- EMBER3D (PyTorch):
- Manuscript: N/A
- Code: https://github.com/kWeissenow/EMBER3D
- HelixFold-single (PaddlePaddle)
- IgFold (PyTorch) - specific to antibody sequences
- Manuscript: https://www.biorxiv.org/content/10.1101/2022.04.20.488972v2
- Code: https://github.com/Graylab/IgFold
- Other: Colab Notebook
- OmegaFold (PyTorch)
- ESM-Fold (PyTorch)
- gget (AF2)
- alphafold_finetune - finetune AlphaFold for Protein-Peptide prediction
- AlphaPulldown - protein-protein interaction screens using AlphaFold-Multimer
- ColabDesign - Backprop through AlphaFold for protein design
- AF2Rank - Rank Decoy Structures/Sequences using AlphaFold
- Manuscript: https://www.biorxiv.org/content/10.1101/2022.03.11.484043v3
- Code: https://github.com/jproney/AF2Rank
- Resource: Colab Notebook
- OpenFold - MSAs for 132K PDBs + 270K UniClust30 predictions for distilation
- MindSpore - MSAs for 570K PDBs + 745K Distillation
- AlphaFold Database (UniRef90 | AF2)
- Manuscript: https://doi.org/10.1093/nar/gkab1061
- Resource: https://alphafold.ebi.ac.uk
- Eukaryotic interactormes (Protein-Protein interactions | RF/AF2)
- Structures of human-transcriptome isoforms (ColabFold/AF2)
- Manuscript: https://doi.org/10.1101/2022.06.08.495354
- Resource: https://www.isoform.io
- AlphaFill - enriching the AlphaFold models with ligands and co-factors (AF2)
- Manuscript: https://www.biorxiv.org/content/10.1101/2021.11.26.470110v1
- Resource: https://alphafill.eu/
- IgFold Database (OAS | IgFold) - predictions specific to antibody sequences
- Lambda PredictProtein (ColabFold|limit:500AA)
- Manuscript: https://doi.org/10.1101/2022.08.04.502750
- Resource: http://embed.predictprotein.org/
- Robetta (RoseTTAFold)
- Resource: https://robetta.bakerlab.org/
- π Moonbear
- Resource: https://www.getmoonbear.com/
- Other: [tweet] Stephanie's announcement