The structural coverage of the human proteome before and after AlphaFold

Abstract

The protein structure field is experiencing a revolution. From the increased throughput of techniques to determine experimental structures, to developments such as cryo-EM that allow us to find the structures of large protein complexes or, more recently, the development of artificial intelligence tools, such as AlphaFold, that can predict with high accuracy the folding of proteins for which the availability of homology templates is limited. Here we quantify the effect of the recently released AlphaFold database of protein structural models in our knowledge on human proteins. Our results indicate that our current baseline for structural coverage of 48%, considering experimentally-derived or template-based homology models, elevates up to 76% when including AlphaFold predictions. At the same time the fraction of dark proteome is reduced from 26% to just 10% when AlphaFold models are considered. Furthermore, although the coverage of disease-associated genes and mutations was near complete before AlphaFold release (69% of Clinvar pathogenic mutations and 88% of oncogenic mutations), AlphaFold models still provide an additional coverage of 3% to 13% of these critically important sets of biomedical genes and mutations. Finally, we show how the contribution of AlphaFold models to the structural coverage of non-human organisms, including important pathogenic bacteria, is significantly larger than that of the human proteome. Overall, our results show that the sequence-structure gap of human proteins has almost disappeared, an outstanding success of direct consequences for the knowledge on the human genome and the derived medical applications.

Dali server: structural unification of protein families

“…Many protein families have diversified beyond recognition by sequence comparison, their evolutionary relationship being recoverable only with support from structure comparison ( 5 ). Recent breakthroughs ( 6 , 7 ) in accurate protein structure prediction dramatically increase the scope of comparative structural studies ( 8 ). Using publically available software ( https://github.com/sokrypton/ColabFold ), structural models can be predicted with confidence for families that do not display any sequence similarity with known protein families.…”

Assessing the clinical utility of protein structural analysis in genomic variant classification: experiences from a diagnostic laboratory

“…The recently released database of structures predicted by the AlphaFold-2 machine-learning method now provides a predicted structure for 98.5% of residues in the human proteome; however, the confidence with which structure can be predicted is highly variable, and only 58% of residues in the AlphaFold database have a predicted Local Distance Difference Test score (pLDDT) of > 70, the lower limit recommended for use in analysis [ 6 , 7 ]. As such, while the AlphaFold database represents a valuable resource for structural analysis, it provides a relatively modest increase in the proportion of the human genome which can be modelled with confidence [ 5 , 8 , 9 ].…”

Characterizing and explaining the impact of disease-associated mutations in proteins without known structures or structural homologs

“…With the current improved state-of-the-art computational structural predictors, like RoseTTAFold and AlphaFold, we can model regions of proteins without any known homologs for about 20% of the residues in humans [ 90 ]. In this study, we modeled the disease-associated human proteins without known homologs [calculated using Basic Local Alignment Search Tool (BLAST)] using RoseTTAFold and AlphaFold.…”