Clustering FunFams using sequence embeddings improves EC purity

Enhancing missense variant pathogenicity prediction with protein language models using VariPred

Computational approaches for predicting the pathogenicity of genetic variants have advanced in recent years. These methods enable researchers to determine the possible clinical impact of rare and novel variants. Historically these prediction methods used hand-crafted features based on structural, evolutionary, or physiochemical properties of the variant. In this study we propose a novel framework that leverages the power of pre-trained protein language models to predict variant pathogenicity. We show that our approach VariPred (Variant impact Predictor) outperforms current state-of-the-art methods by using an end-to-end model that only requires the protein sequence as input. Using one of the best-performing protein language models (ESM-1b), we establish a robust classifier that requires no calculation of structural features or multiple sequence alignments. We compare the performance of VariPred with other representative models including 3Cnet, Polyphen-2, REVEL, MetaLR, FATHMM and ESM variant. VariPred performs as well as, or in most cases better than these other predictors using six variant impact prediction benchmarks despite requiring only sequence data and no pre-processing of the data.

SAP: Synteny-aware gene function prediction for bacteria using protein embeddings

Motivation

Today, we know the function of only a small fraction of the protein sequences predicted from genomic data. This problem is even more salient for bacteria, which represent some of the most phylogenetically and metabolically diverse taxa on Earth. This low rate of bacterial gene annotation is compounded by the fact that most function prediction algorithms have focused on eukaryotes, and conventional annotation approaches rely on the presence of similar sequences in existing databases. However, often there are no such sequences for novel bacterial proteins. Thus, we need improved gene function prediction methods tailored for prokaryotes. Recently, transformer-based language models - adopted from the natural language processing field - have been used to obtain new representations of proteins, to replace amino acid sequences. These representations, referred to as protein embeddings, have shown promise for improving annotation of eukaryotes, but there have been only limited applications on bacterial genomes.

Results

To predict gene functions in bacteria, we developed SAP, a novel synteny-aware gene function prediction tool based on protein embeddings from state-of-the-art protein language models. SAP also leverages the unique operon structure of bacteria through conserved synteny. SAP outperformed both conventional sequence-based annotation methods and state-of-the-art methods on multiple bacterial species, including for distant homolog detection, where the sequence similarity to the proteins in the training set was as low as 40%. Using SAP to identify gene functions across diverse enterococci, of which some species are major clinical threats, we identified 11 previously unrecognized putative novel toxins, with potential significance to human and animal health.

Availability

https://github.com/AbeelLab/sap.

Contact

t.abeel@tudelft.nl.

Supplementary information

Supplementary data are available at Bioinformatics online.

Novel machine learning approaches revolutionize protein knowledge

Breakthrough methods in machine learning (ML), protein structure prediction, and novel ultrafast structural aligners are revolutionizing structural biology. Obtaining accurate models of proteins and annotating their functions on a large scale is no longer limited by time and resources. The most recent method to be top ranked by the Critical Assessment of Structure Prediction (CASP) assessment, AlphaFold 2 (AF2), is capable of building structural models with an accuracy comparable to that of experimental structures. Annotations of 3D models are keeping pace with the deposition of the structures due to advancements in protein language models (pLMs) and structural aligners that help validate these transferred annotations. In this review we describe how recent developments in ML for protein science are making large-scale structural bioinformatics available to the general scientific community.

AlphaFold2 reveals commonalities and novelties in protein structure space for 21 model organisms

Deep-learning (DL) methods like DeepMind's AlphaFold2 (AF2) have led to substantial improvements in protein structure prediction. We analyse confident AF2 models from 21 model organisms using a new classification protocol (CATH-Assign) which exploits novel DL methods for structural comparison and classification. Of ~370,000 confident models, 92% can be assigned to 3253 superfamilies in our CATH domain superfamily classification. The remaining cluster into 2367 putative novel superfamilies. Detailed manual analysis on 618 of these, having at least one human relative, reveal extremely remote homologies and further unusual features. Only 25 novel superfamilies could be confirmed. Although most models map to existing superfamilies, AF2 domains expand CATH by 67% and increases the number of unique 'global' folds by 36% and will provide valuable insights on structure function relationships. CATH-Assign will harness the huge expansion in structural data provided by DeepMind to rationalise evolutionary changes driving functional divergence.

KinFams: De-Novo Classification of Protein Kinases Using CATH Functional Units

Protein kinases are important targets for treating human disorders, and they are the second most targeted families after G-protein coupled receptors. Several resources provide classification of kinases into evolutionary families (based on sequence homology); however, very few systematically classify functional families (FunFams) comprising evolutionary relatives that share similar functional properties. We have developed the FunFam-MARC (Multidomain ARchitecture-based Clustering) protocol, which uses multi-domain architectures of protein kinases and specificity-determining residues for functional family classification. FunFam-MARC predicts 2210 kinase functional families (KinFams), which have increased functional coherence, in terms of EC annotations, compared to the widely used KinBase classification. Our protocol provides a comprehensive classification for kinase sequences from >10,000 organisms. We associate human KinFams with diseases and drugs and identify 28 druggable human KinFams, i.e., enriched in clinically approved drugs. Since relatives in the same druggable KinFam tend to be structurally conserved, including the drug-binding site, these KinFams may be valuable for shortlisting therapeutic targets. Information on the human KinFams and associated 3D structures from AlphaFold2 are provided via our CATH FTP website and Zenodo. This gives the domain structure representative of each KinFam together with information on any drug compounds available. For 32% of the KinFams, we provide information on highly conserved residue sites that may be associated with specificity.

LambdaPP: Fast and accessible protein-specific phenotype predictions

The availability of accurate and fast artificial intelligence (AI) solutions predicting aspects of proteins are revolutionizing experimental and computational molecular biology. The webserver LambdaPP aspires to supersede PredictProtein, the first internet server making AI protein predictions available in 1992. Given a protein sequence as input, LambdaPP provides easily accessible visualizations of protein 3D structure, along with predictions at the protein level (GeneOntology, subcellular location), and the residue level (binding to metal ions, small molecules, and nucleotides; conservation; intrinsic disorder; secondary structure; alpha-helical and beta-barrel transmembrane segments; signal-peptides; variant effect) in seconds. The structure prediction provided by LambdaPP-leveraging ColabFold and computed in minutes-is based on MMseqs2 multiple sequence alignments. All other feature prediction methods are based on the pLM ProtT5. Queried by a protein sequence, LambdaPP computes protein and residue predictions almost instantly for various phenotypes, including 3D structure and aspects of protein function. LambdaPP is freely available for everyone to use under embed.predictprotein.org, the interactive results for the case study can be found under https://embed.predictprotein.org/o/Q9NZC2. The frontend of LambdaPP can be found on GitHub (github.com/sacdallago/embed.predictprotein.org), and can be freely used and distributed under the academic free use license (AFL-2). For high-throughput applications, all methods can be executed locally via the bio-embeddings (bioembeddings.com) python package, or docker image at ghcr.io/bioembeddings/bio_embeddings, which also includes the backend of LambdaPP.

General strategies for using amino acid sequence data to guide biochemical investigation of protein function

The rapid increase of '-omics' data warrants the reconsideration of experimental strategies to investigate general protein function. Studying individual members of a protein family is likely insufficient to provide a complete mechanistic understanding of family functions, especially for diverse families with thousands of known members. Strategies that exploit large amounts of available amino acid sequence data can inspire and guide biochemical experiments, generating broadly applicable insights into a given family. Here we review several methods that utilize abundant sequence data to focus experimental efforts and identify features truly representative of a protein family or domain. First, coevolutionary relationships between residues within primary sequences can be successfully exploited to identify structurally and/or functionally important positions for experimental investigation. Second, functionally important variable residue positions typically occupy a limited sequence space, a property useful for guiding biochemical characterization of the effects of the most physiologically and evolutionarily relevant amino acids. Third, amino acid sequence variation within domains shared between different protein families can be used to sort a particular domain into multiple subtypes, inspiring further experimental designs. Although generally applicable to any kind of protein domain because they depend solely on amino acid sequences, the second and third approaches are reviewed in detail because they appear to have been used infrequently and offer immediate opportunities for new advances. Finally, we speculate that future technologies capable of analyzing and manipulating conserved and variable aspects of the three-dimensional structures of a protein family could lead to broad insights not attainable by current methods.

Nearest neighbor search on embeddings rapidly identifies distant protein relations

Since 1992, all state-of-the-art methods for fast and sensitive identification of evolutionary, structural, and functional relations between proteins (also referred to as "homology detection") use sequences and sequence-profiles (PSSMs). Protein Language Models (pLMs) generalize sequences, possibly capturing the same constraints as PSSMs, e.g., through embeddings. Here, we explored how to use such embeddings for nearest neighbor searches to identify relations between protein pairs with diverged sequences (remote homology detection for levels of <20% pairwise sequence identity, PIDE). While this approach excelled for proteins with single domains, we demonstrated the current challenges applying this to multi-domain proteins and presented some ideas how to overcome existing limitations, in principle. We observed that sufficiently challenging data set separations were crucial to provide deeply relevant insights into the behavior of nearest neighbor search when applied to the protein embedding space, and made all our methods readily available for others.

SETH predicts nuances of residue disorder from protein embeddings

Predictions for millions of protein three-dimensional structures are only a few clicks away since the release of AlphaFold2 results for UniProt. However, many proteins have so-called intrinsically disordered regions (IDRs) that do not adopt unique structures in isolation. These IDRs are associated with several diseases, including Alzheimer’s Disease. We showed that three recent disorder measures of AlphaFold2 predictions (pLDDT, “experimentally resolved” prediction and “relative solvent accessibility”) correlated to some extent with IDRs. However, expert methods predict IDRs more reliably by combining complex machine learning models with expert-crafted input features and evolutionary information from multiple sequence alignments (MSAs). MSAs are not always available, especially for IDRs, and are computationally expensive to generate, limiting the scalability of the associated tools. Here, we present the novel method SETH that predicts residue disorder from embeddings generated by the protein Language Model ProtT5, which explicitly only uses single sequences as input. Thereby, our method, relying on a relatively shallow convolutional neural network, outperformed much more complex solutions while being much faster, allowing to create predictions for the human proteome in about 1 hour on a consumer-grade PC with one NVIDIA GeForce RTX 3060. Trained on a continuous disorder scale (CheZOD scores), our method captured subtle variations in disorder, thereby providing important information beyond the binary classification of most methods. High performance paired with speed revealed that SETH’s nuanced disorder predictions for entire proteomes capture aspects of the evolution of organisms. Additionally, SETH could also be used to filter out regions or proteins with probable low-quality AlphaFold2 3D structures to prioritize running the compute-intensive predictions for large data sets. SETH is freely publicly available at: https://github.com/Rostlab/SETH.

Embeddings from protein language models predict conservation and variant effects

The emergence of SARS-CoV-2 variants stressed the demand for tools allowing to interpret the effect of single amino acid variants (SAVs) on protein function. While Deep Mutational Scanning (DMS) sets continue to expand our understanding of the mutational landscape of single proteins, the results continue to challenge analyses. Protein Language Models (pLMs) use the latest deep learning (DL) algorithms to leverage growing databases of protein sequences. These methods learn to predict missing or masked amino acids from the context of entire sequence regions. Here, we used pLM representations (embeddings) to predict sequence conservation and SAV effects without multiple sequence alignments (MSAs). Embeddings alone predicted residue conservation almost as accurately from single sequences as ConSeq using MSAs (two-state Matthews Correlation Coefficient-MCC-for ProtT5 embeddings of 0.596 ± 0.006 vs. 0.608 ± 0.006 for ConSeq). Inputting the conservation prediction along with BLOSUM62 substitution scores and pLM mask reconstruction probabilities into a simplistic logistic regression (LR) ensemble for Variant Effect Score Prediction without Alignments (VESPA) predicted SAV effect magnitude without any optimization on DMS data. Comparing predictions for a standard set of 39 DMS experiments to other methods (incl. ESM-1v, DeepSequence, and GEMME) revealed our approach as competitive with the state-of-the-art (SOTA) methods using MSA input. No method outperformed all others, neither consistently nor statistically significantly, independently of the performance measure applied (Spearman and Pearson correlation). Finally, we investigated binary effect predictions on DMS experiments for four human proteins. Overall, embedding-based methods have become competitive with methods relying on MSAs for SAV effect prediction at a fraction of the costs in computing/energy. Our method predicted SAV effects for the entire human proteome (~ 20 k proteins) within 40 min on one Nvidia Quadro RTX 8000. All methods and data sets are freely available for local and online execution through bioembeddings.com, https://github.com/Rostlab/VESPA , and PredictProtein.

Organizing the bacterial annotation space with amino acid sequence embeddings

BACKGROUND

Due to the ever-expanding gap between the number of proteins being discovered and their functional characterization, protein function inference remains a fundamental challenge in computational biology. Currently, known protein annotations are organized in human-curated ontologies, however, all possible protein functions may not be organized accurately. Meanwhile, recent advancements in natural language processing and machine learning have developed models which embed amino acid sequences as vectors in n-dimensional space. So far, these embeddings have primarily been used to classify protein sequences using manually constructed protein classification schemes.

RESULTS

In this work, we describe the use of amino acid sequence embeddings as a systematic framework for studying protein ontologies. Using a sequence embedding, we show that the bacterial carbohydrate metabolism class within the SEED annotation system contains 48 clusters of embedded sequences despite this class containing 29 functional labels. Furthermore, by embedding Bacillus amino acid sequences with unknown functions, we show that these unknown sequences form clusters that are likely to have similar biological roles.

CONCLUSIONS

This study demonstrates that amino acid sequence embeddings may be a powerful tool for developing more robust ontologies for annotating protein sequence data. In addition, embeddings may be beneficial for clustering protein sequences with unknown functions and selecting optimal candidate proteins to characterize experimentally.

A roadmap for the functional annotation of protein families: a community perspective

Over the last 25 years, biology has entered the genomic era and is becoming a science of ‘big data’. Most interpretations of genomic analyses rely on accurate functional annotations of the proteins encoded by more than 500 000 genomes sequenced to date. By different estimates, only half the predicted sequenced proteins carry an accurate functional annotation, and this percentage varies drastically between different organismal lineages. Such a large gap in knowledge hampers all aspects of biological enterprise and, thereby, is standing in the way of genomic biology reaching its full potential. A brainstorming meeting to address this issue funded by the National Science Foundation was held during 3–4 February 2022. Bringing together data scientists, biocurators, computational biologists and experimentalists within the same venue allowed for a comprehensive assessment of the current state of functional annotations of protein families. Further, major issues that were obstructing the field were identified and discussed, which ultimately allowed for the proposal of solutions on how to move forward.

Contrastive learning on protein embeddings enlightens midnight zone

Experimental structures are leveraged through multiple sequence alignments, or more generally through homology-based inference (HBI), facilitating the transfer of information from a protein with known annotation to a query without any annotation. A recent alternative expands the concept of HBI from sequence-distance lookup to embedding-based annotation transfer (EAT). These embeddings are derived from protein Language Models (pLMs). Here, we introduce using single protein representations from pLMs for contrastive learning. This learning procedure creates a new set of embeddings that optimizes constraints captured by hierarchical classifications of protein 3D structures defined by the CATH resource. The approach, dubbed ProtTucker, has an improved ability to recognize distant homologous relationships than more traditional techniques such as threading or fold recognition. Thus, these embeddings have allowed sequence comparison to step into the 'midnight zone' of protein similarity, i.e. the region in which distantly related sequences have a seemingly random pairwise sequence similarity. The novelty of this work is in the particular combination of tools and sampling techniques that ascertained good performance comparable or better to existing state-of-the-art sequence comparison methods. Additionally, since this method does not need to generate alignments it is also orders of magnitudes faster. The code is available at https://github.com/Rostlab/EAT.