Highly accurate protein structure prediction for the human proteome

By Kathryn Tunyasuvunakool, Jonas Adler, Zachary Wu, Tim Green, Michal Zielinski, Augustin Žídek, Alex Bridgland, Andrew Cowie, Clemens Meyer, Agata Laydon and 23 more. 2021

2,524 citations recorded by the source corpus.

Protein structures can provide invaluable information, both for reasoning about biological processes and for enabling interventions such as structure-based drug development or targeted mutagenesis. After decades of effort, 17% of the total residues in human protein sequences are covered by an experimentally determined structure1. Here we markedly expand the structural coverage of the proteome by applying the state-of-the-art machine learning method, AlphaFold2, at a scale that covers almost the entire human proteome (98.5% of human proteins). The resulting dataset covers 58% of residues with a confident prediction, of which a subset (36% of all residues) have very high confidence. We introduce several metrics developed by building on the AlphaFold model and use them to interpret the dataset, identifying strong multi-domain predictions as well as regions that are likely to be disordered. Finally, we provide some case studies to illustrate how high-quality predictions could be used to generate biological hypotheses. We are making our predictions freely available to the community and anticipate that routine large-scale and high-accuracy structure prediction will become an important tool that will allow new questions to be addressed from a structural perspective. AlphaFold is used to predict the structures of almost all of the proteins in the human proteome—the availability of high-confidence predicted structures could enable new avenues of investigation from a structural perspective.

Subfields: Molecular Biology, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Computational Biology & Bioinformatics

Citations in the corpus

State-specific protein–ligand complex structure prediction with a multiscale deep generative model (2022, positive): “…on a standard GPU once all auxiliary features are gathered, thereby enabling the exploration of protein–ligand interactions on the vast space of sequence and chemical spaces by leveraging rapidly evolving databases of computationally modelled protein structures 9,54 and chemical substances 55,56 .”
Sampling alternative conformational states of transporters and receptors with AlphaFold2 (2022, positive): “A performance breakthrough was unveiled during CASP14 by AlphaFold2 (AF2) ( Jumper et al., 2021a ; Tunyasuvunakool et al., 2021 ; Pereira et al., 2021 ), which achieved remarkably accurate de novo structure prediction.”
Highly accurate protein structure prediction with AlphaFold (2021, background): “This opens up the exciting possibility of predicting structures at the proteome-scale and beyond—in a companion paper 39 , we demonstrate the application of AlphaFold to the entire human proteome 39 .”
ColabFold: making protein folding accessible to all (2022, methodology): “Concurrently, Tunyasuvunakool et al. 9 developed an AlphaFold2 Jupyter Notebook for Google Colaboratory (referred to as AlphaFold-Colab), for which the input MSA is built by searching with HMMer against the UniProt Reference Clusters (UniRef90) and an eightfold-reduced environmental database.”
AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models (2021, methodology): “The methodology of AlphaFold and insights gained from the predictions for the complete human proteome have been described recently (9,13).”
Scaffolding protein functional sites using deep learning (2022, background): “S20; Table S3; Supplementary Text) (41), and our designed structures are predicted more confidently from their (single) sequences than most native proteins with known crystal structures, and on par with structurally validated de novo designed proteins (Fig.”
Biasing AlphaFold2 to predict GPCRs and kinases with user-defined functional or structural properties (2022, background): “AF2 makes it straightforward to predict a protein structure from a protein sequence and has provided millions of protein models with estimated accuracy (Tunyasuvunakool et al., 2021).”
Protein language model embeddings for fast, accurate, alignment-free protein structure prediction (2021, background): “One solution is to generate multiple MSAs with different parameters, alignment tools and databases (Jain et al., 2021; Zhang et al., 2020), rendering the input generation even more time-consuming, e.g. 212 minutes for our test sets of 31 proteins on an Intel Xeon Gold 6248 (Results).”
OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization (2023, positive): “It has also been used to predict the structures of hundreds of millions of proteins 19–21 .”
Computational approaches streamlining drug discovery (2023, background): “99,100) or RosettaFold 101 3D models, which already show utility in many applications, including protein–protein and protein–peptide docking 102 .”
Impact of protein conformational diversity on AlphaFold predictions (2021, background): “Far from disappointing, this observation was expected since a large set of redundant protein structures (conformers) would have been required during the neural network training processes [5] in order to predict conformational diversity.”
Clustering predicted structures at the scale of the known protein universe (2023, background): “For example, protein family analysis of the initial release of about 365,000 structures 10,16 , covering the proteomes of humans and 20 model organisms, suggested that 92% of predicted domains within this set match existing domain superfamilies.”