SIFTS: Structure Integration with Function, Taxonomy and Sequences resource

CASTpFold: Computed Atlas of Surface Topography of the universe of protein Folds

Geometric and topological properties of protein structures, including surface pockets, interior cavities and cross channels, are of fundamental importance for proteins to carry out their functions. Computed Atlas of Surface Topography of proteins (CASTp) is a widely used web server for locating, delineating, and measuring these geometric and topological properties of protein structures. Recent developments in AI-based protein structure prediction such as AlphaFold2 (AF2) have significantly expanded our knowledge on protein structures. Here we present CASTpFold, a continuation of CASTp that provides accurate and comprehensive identifications and quantifications of protein topography. It now provides (i) results on an expanded database of proteins, including the Protein Data Bank (PDB) and non-singleton representative structures of AlphaFold2 structures, covering 183 million AF2 structures; (ii) functional pockets prediction with corresponding Gene Ontology (GO) terms or Enzyme Commission (EC) numbers for AF2-predicted structures and (iii) pocket similarity search function for surface and protein-protein interface pockets. The CASTpFold web server is freely accessible at https://cfold.bme.uic.edu/castpfold/.

PPInterface: A Comprehensive Dataset of 3D Protein-Protein Interface Structures

The PPInterface dataset contains 815,082 interface structures, providing the most comprehensive structural information on protein-protein interfaces. This resource is extracted from over 215,000 three-dimensional protein structures stored in the Protein Data Bank (PDB). The dataset contains a wide range of protein complexes, providing a wealth of information for researchers investigating the structural properties of protein-protein interactions. The accompanying web server has a user-friendly interface that allows for efficient search and download functions. Researchers can access detailed information on protein interface structures, visualize them, and explore a variety of features, increasing the dataset's utility and accessibility. The dataset and web server can be found at https://3dpath.ku.edu.tr/PPInt/.

PDB NextGen Archive: centralizing access to integrated annotations and enriched structural information by the Worldwide Protein Data Bank

The Protein Data Bank (PDB) is the global repository for public-domain experimentally determined 3D biomolecular structural information. The archival nature of the PDB presents certain challenges pertaining to updating or adding associated annotations from trusted external biodata resources. While each Worldwide PDB (wwPDB) partner has made best efforts to provide up-to-date external annotations, accessing and integrating information from disparate wwPDB data centers can be an involved process. To address this issue, the wwPDB has established the PDB Next Generation (or NextGen) Archive, developed to centralize and streamline access to enriched structural annotations from wwPDB partners and trusted external sources. At present, the NextGen Archive provides mappings between experimentally determined 3D structures of proteins and UniProt amino acid sequences, domain annotations from Pfam, SCOP2 and CATH databases and intra-molecular connectivity information. Since launch, the PDB NextGen Archive has seen substantial user engagement with over 3.5 million data file downloads, ensuring researchers have access to accurate, up-to-date and easily accessible structural annotations. Database URL: http://www.wwpdb.org/ftp/pdb-nextgen-archive-site.

Decoding glycosylation potential from protein structure across human glycoproteins with a multi-view recurrent neural network

Glycosylation is described as a non-templated biosynthesis. Yet, the template-free premise is antithetical to the observation that different N-glycans are consistently placed at specific sites. It has been proposed that glycosite-proximal protein structures could constrain glycosylation and explain the observed microheterogeneity. Using site-specific glycosylation data, we trained a hybrid neural network to parse glycosites (recurrent neural network) and match them to feasible N-glycosylation events (graph neural network). From glycosite-flanking sequences, the algorithm predicts most human N-glycosylation events documented in the GlyConnect database and proposed structures corresponding to observed monosaccharide composition of the glycans at these sites. The algorithm also recapitulated glycosylation in Enhanced Aromatic Sequons, SARS-CoV-2 spike, and IgG3 variants, thus demonstrating the ability of the algorithm to predict both glycan structure and abundance. Thus, protein structure constrains glycosylation, and the neural network enables predictive in silico glycosylation of uncharacterized or novel protein sequences and genetic variants.

ZEPPI: Proteome-scale sequence-based evaluation of protein-protein interaction models

We introduce ZEPPI (Z-score Evaluation of Protein-Protein Interfaces), a framework to evaluate structural models of a complex based on sequence coevolution and conservation involving residues in protein-protein interfaces. The ZEPPI score is calculated by comparing metrics for an interface to those obtained from randomly chosen residues. Since contacting residues are defined by the structural model, this obviates the need to account for indirect interactions. Further, although ZEPPI relies on species-paired multiple sequence alignments, its focus on interfacial residues allows it to leverage quite shallow alignments. ZEPPI can be implemented on a proteome-wide scale and is applied here to millions of structural models of dimeric complexes in the Escherichia coli and human interactomes found in the PrePPI database. PrePPI's scoring function is based primarily on the evaluation of protein-protein interfaces, and ZEPPI adds a new feature to this analysis through the incorporation of evolutionary information. ZEPPI performance is evaluated through applications to experimentally determined complexes and to decoys from the CASP-CAPRI experiment. As we discuss, the standard CAPRI scores used to evaluate docking models are based on model quality and not on the ability to give yes/no answers as to whether two proteins interact. ZEPPI is able to detect weak signals from PPI models that the CAPRI scores define as incorrect and, similarly, to identify potential PPIs defined as low confidence by the current PrePPI scoring function. A number of examples that illustrate how the combination of PrePPI and ZEPPI can yield functional hypotheses are provided.

CASTpFold: Computed Atlas of Surface Topography of the universe of protein Folds

Geometric and topological properties of protein structures, including surface pockets, interior cavities, and cross channels, are of fundamental importance for proteins to carry out their functions. Computed Atlas of Surface Topography of proteins (CASTp) is a widely used web server for locating, delineating, and measuring these geometric and topological properties of protein structures. Recent developments in AI-based protein structure prediction such as AlphaFold2 (AF2) have significantly expanded our knowledge on protein structures. Here we present CASTpFold, a continuation of CASTp that provides accurate and comprehensive identifications and quantifications of protein topography. It now provides (i) results on an expanded database of proteins, including the Protein Data Bank (PDB) and non-singleton representative structures of AlphaFold2 structures, covering 183 million AF2 structures; (ii) functional pockets prediction with corresponding Gene Ontology (GO) terms or Enzyme Commission (EC) numbers for AF2-predicted structures; and (iii) pocket similarity search function for surface and protein-protein interface pockets. The CASTpFold web server is freely accessible at https://cfold.bme.uic.edu/castpfold/.

Sliding Window INteraction Grammar (SWING): a generalized interaction language model for peptide and protein interactions

The explosion of sequence data has allowed the rapid growth of protein language models (pLMs). pLMs have now been employed in many frameworks including variant-effect and peptide-specificity prediction. Traditionally, for protein-protein or peptide-protein interactions (PPIs), corresponding sequences are either co-embedded followed by post-hoc integration or the sequences are concatenated prior to embedding. Interestingly, no method utilizes a language representation of the interaction itself. We developed an interaction LM (iLM), which uses a novel language to represent interactions between protein/peptide sequences. Sliding Window Interaction Grammar (SWING) leverages differences in amino acid properties to generate an interaction vocabulary. This vocabulary is the input into a LM followed by a supervised prediction step where the LM's representations are used as features. SWING was first applied to predicting peptide:MHC (pMHC) interactions. SWING was not only successful at generating Class I and Class II models that have comparable prediction to state-of-the-art approaches, but the unique Mixed Class model was also successful at jointly predicting both classes. Further, the SWING model trained only on Class I alleles was predictive for Class II, a complex prediction task not attempted by any existing approach. For de novo data, using only Class I or Class II data, SWING also accurately predicted Class II pMHC interactions in murine models of SLE (MRL/lpr model) and T1D (NOD model), that were validated experimentally. To further evaluate SWING's generalizability, we tested its ability to predict the disruption of specific protein-protein interactions by missense mutations. Although modern methods like AlphaMissense and ESM1b can predict interfaces and variant effects/pathogenicity per mutation, they are unable to predict interaction-specific disruptions. SWING was successful at accurately predicting the impact of both Mendelian mutations and population variants on PPIs. This is the first generalizable approach that can accurately predict interaction-specific disruptions by missense mutations with only sequence information. Overall, SWING is a first-in-class generalizable zero-shot iLM that learns the language of PPIs.

Identifying protein conformational states in the Protein Data Bank: Toward unlocking the potential of integrative dynamics studies

Studying protein dynamics and conformational heterogeneity is crucial for understanding biomolecular systems and treating disease. Despite the deposition of over 215 000 macromolecular structures in the Protein Data Bank and the advent of AI-based structure prediction tools such as AlphaFold2, RoseTTAFold, and ESMFold, static representations are typically produced, which fail to fully capture macromolecular motion. Here, we discuss the importance of integrating experimental structures with computational clustering to explore the conformational landscapes that manifest protein function. We describe the method developed by the Protein Data Bank in Europe - Knowledge Base to identify distinct conformational states, demonstrate the resource's primary use cases, through examples, and discuss the need for further efforts to annotate protein conformations with functional information. Such initiatives will be crucial in unlocking the potential of protein dynamics data, expediting drug discovery research, and deepening our understanding of macromolecular mechanisms.

A unified analysis of evolutionary and population constraint in protein domains highlights structural features and pathogenic sites

Protein evolution is constrained by structure and function, creating patterns in residue conservation that are routinely exploited to predict structure and other features. Similar constraints should affect variation across individuals, but it is only with the growth of human population sequencing that this has been tested at scale. Now, human population constraint has established applications in pathogenicity prediction, but it has not yet been explored for structural inference. Here, we map 2.4 million population variants to 5885 protein families and quantify residue-level constraint with a new Missense Enrichment Score (MES). Analysis of 61,214 structures from the PDB spanning 3661 families shows that missense depleted sites are enriched in buried residues or those involved in small-molecule or protein binding. MES is complementary to evolutionary conservation and a combined analysis allows a new classification of residues according to a conservation plane. This approach finds functional residues that are evolutionarily diverse, which can be related to specificity, as well as family-wide conserved sites that are critical for folding or function. We also find a possible contrast between lethal and non-lethal pathogenic sites, and a surprising clinical variant hot spot at a subset of missense enriched positions.

Classifying protein kinase conformations with machine learning

Protein kinases are key actors of signaling networks and important drug targets. They cycle between active and inactive conformations, distinguished by a few elements within the catalytic domain. One is the activation loop, whose conserved DFG motif can occupy DFG-in, DFG-out, and some rarer conformations. Annotation and classification of the structural kinome are important, as different conformations can be targeted by different inhibitors and activators. Valuable resources exist; however, large-scale applications will benefit from increased automation and interpretability of structural annotation. Interpretable machine learning models are described for this purpose, based on ensembles of decision trees. To train them, a set of catalytic domain sequences and structures was collected, somewhat larger and more diverse than existing resources. The structures were clustered based on the DFG conformation and manually annotated. They were then used as training input. Two main models were constructed, which distinguished active/inactive and in/out/other DFG conformations. They considered initially 1692 structural variables, spanning the whole catalytic domain, then identified ("learned") a small subset that sufficed for accurate classification. The first model correctly labeled all but 3 of 3289 structures as active or inactive, while the second assigned the correct DFG label to all but 17 of 8826 structures. The most potent classifying variables were all related to well-known structural elements in or near the activation loop and their ranking gives insights into the conformational preferences. The models were used to automatically annotate 3850 kinase structures predicted recently with the Alphafold2 tool, showing that Alphafold2 reproduced the active/inactive but not the DFG-in proportions seen in the Protein Data Bank. We expect the models will be useful for understanding and engineering kinases.

Structurally-informed human interactome reveals proteome-wide perturbations by disease mutations

Human genome sequencing studies have identified numerous loci associated with complex diseases. However, translating human genetic and genomic findings to disease pathobiology and therapeutic discovery remains a major challenge at multiscale interactome network levels. Here, we present a deep-learning-based ensemble framework, termed PIONEER (Protein-protein InteractiOn iNtErfacE pRediction), that accurately predicts protein binding partner-specific interfaces for all known protein interactions in humans and seven other common model organisms, generating comprehensive structurally-informed protein interactomes. We demonstrate that PIONEER outperforms existing state-of-the-art methods. We further systematically validated PIONEER predictions experimentally through generating 2,395 mutations and testing their impact on 6,754 mutation-interaction pairs, confirming the high quality and validity of PIONEER predictions. We show that disease-associated mutations are enriched in PIONEER-predicted protein-protein interfaces after mapping mutations from ~60,000 germline exomes and ~36,000 somatic genomes. We identify 586 significant protein-protein interactions (PPIs) enriched with PIONEER-predicted interface somatic mutations (termed oncoPPIs) from pan-cancer analysis of ~11,000 tumor whole-exomes across 33 cancer types. We show that PIONEER-predicted oncoPPIs are significantly associated with patient survival and drug responses from both cancer cell lines and patient-derived xenograft mouse models. We identify a landscape of PPI-perturbing tumor alleles upon ubiquitination by E3 ligases, and we experimentally validate the tumorigenic KEAP1-NRF2 interface mutation p.Thr80Lys in non-small cell lung cancer. We show that PIONEER-predicted PPI-perturbing alleles alter protein abundance and correlates with drug responses and patient survival in colon and uterine cancers as demonstrated by proteogenomic data from the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium. PIONEER, implemented as both a web server platform and a software package, identifies functional consequences of disease-associated alleles and offers a deep learning tool for precision medicine at multiscale interactome network levels.

ARCTIC-3D: automatic retrieval and clustering of interfaces in complexes from 3D structural information

The formation of a stable complex between proteins lies at the core of a wide variety of biological processes and has been the focus of countless experiments. The huge amount of information contained in the protein structural interactome in the Protein Data Bank can now be used to characterise and classify the existing biological interfaces. We here introduce ARCTIC-3D, a fast and user-friendly data mining and clustering software to retrieve data and rationalise the interface information associated with the protein input data. We demonstrate its use by various examples ranging from showing the increased interaction complexity of eukaryotic proteins, 20% of which on average have more than 3 different interfaces compared to only 10% for prokaryotes, to associating different functions to different interfaces. In the context of modelling biomolecular assemblies, we introduce the concept of "recognition entropy", related to the number of possible interfaces of the components of a protein-protein complex, which we demonstrate to correlate with the modelling difficulty in classical docking approaches. The identified interface clusters can also be used to generate various combinations of interface-specific restraints for integrative modelling. The ARCTIC-3D software is freely available at github.com/haddocking/arctic3d and can be accessed as a web-service at wenmr.science.uu.nl/arctic3d.

Integration of pre-trained protein language models into geometric deep learning networks

Geometric deep learning has recently achieved great success in non-Euclidean domains, and learning on 3D structures of large biomolecules is emerging as a distinct research area. However, its efficacy is largely constrained due to the limited quantity of structural data. Meanwhile, protein language models trained on substantial 1D sequences have shown burgeoning capabilities with scale in a broad range of applications. Several preceding studies consider combining these different protein modalities to promote the representation power of geometric neural networks but fail to present a comprehensive understanding of their benefits. In this work, we integrate the knowledge learned by well-trained protein language models into several state-of-the-art geometric networks and evaluate a variety of protein representation learning benchmarks, including protein-protein interface prediction, model quality assessment, protein-protein rigid-body docking, and binding affinity prediction. Our findings show an overall improvement of 20% over baselines. Strong evidence indicates that the incorporation of protein language models' knowledge enhances geometric networks' capacity by a significant margin and can be generalized to complex tasks.

GPCRome-wide analysis of G-protein-coupling diversity using a computational biology approach

GPCRs are master regulators of cell signaling by transducing extracellular stimuli into the cell via selective coupling to intracellular G-proteins. Here we present a computational analysis of the structural determinants of G-protein-coupling repertoire of experimental and predicted 3D GPCR-G-protein complexes. Interface contact analysis recapitulates structural hallmarks associated with G-protein-coupling specificity, including TM5, TM6 and ICLs. We employ interface contacts as fingerprints to cluster G_s vs G_i complexes in an unsupervised fashion, suggesting that interface residues contribute to selective coupling. We experimentally confirm on a promiscuous receptor (CCKAR) that mutations of some of these specificity-determining positions bias the coupling selectivity. Interestingly, G_s-GPCR complexes have more conserved interfaces, while G_i/o proteins adopt a wider number of alternative docking poses, as assessed via structural alignments of representative 3D complexes. Binding energy calculations demonstrate that distinct structural properties of the complexes are associated to higher stability of G_s than G_i/o complexes. AlphaFold2 predictions of experimental binary complexes confirm several of these structural features and allow us to augment the structural coverage of poorly characterized complexes such as G_12/13.

Effective holistic characterization of small molecule effects using heterogeneous biological networks

The two most common reasons for attrition in therapeutic clinical trials are efficacy and safety. We integrated heterogeneous data to create a human interactome network to comprehensively describe drug behavior in biological systems, with the goal of accurate therapeutic candidate generation. The Computational Analysis of Novel Drug Opportunities (CANDO) platform for shotgun multiscale therapeutic discovery, repurposing, and design was enhanced by integrating drug side effects, protein pathways, protein-protein interactions, protein-disease associations, and the Gene Ontology, and complemented with its existing drug/compound, protein, and indication libraries. These integrated networks were reduced to a "multiscale interactomic signature" for each compound that describe its functional behavior as vectors of real values. These signatures are then used for relating compounds to each other with the hypothesis that similar signatures yield similar behavior. Our results indicated that there is significant biological information captured within our networks (particularly via side effects) which enhance the performance of our platform, as evaluated by performing all-against-all leave-one-out drug-indication association benchmarking as well as generating novel drug candidates for colon cancer and migraine disorders corroborated via literature search. Further, drug impacts on pathways derived from computed compound-protein interaction scores served as the features for a random forest machine learning model trained to predict drug-indication associations, with applications to mental disorders and cancer metastasis highlighted. This interactomic pipeline highlights the ability of Computational Analysis of Novel Drug Opportunities to accurately relate drugs in a multitarget and multiscale context, particularly for generating putative drug candidates using the information gleaned from indirect data such as side effect profiles and protein pathway information.

Unified access to up-to-date residue-level annotations from UniProtKB and other biological databases for PDB data

More than 61,000 proteins have up-to-date correspondence between their amino acid sequence (UniProtKB) and their 3D structures (PDB), enabled by the Structure Integration with Function, Taxonomy and Sequences (SIFTS) resource. SIFTS incorporates residue-level annotations from many other biological resources. SIFTS data is available in various formats like XML, CSV and TSV format or also accessible via the PDBe REST API but always maintained separately from the structure data (PDBx/mmCIF file) in the PDB archive. Here, we extended the wwPDB PDBx/mmCIF data dictionary with additional categories to accommodate SIFTS data and added the UniProtKB, Pfam, SCOP2, and CATH residue-level annotations directly into the PDBx/mmCIF files from the PDB archive. With the integrated UniProtKB annotations, these files now provide consistent numbering of residues in different PDB entries allowing easy comparison of structure models. The extended dictionary yields a more consistent, standardised metadata description without altering the core PDB information. This development enables up-to-date cross-reference information at the residue level resulting in better data interoperability, supporting improved data analysis and visualisation.

PrePCI: A structure- and chemical similarity-informed database of predicted protein compound interactions

We describe the Predicting Protein-Compound Interactions (PrePCI) database which comprises over 5 billion predicted interactions between 6.8 million chemical compounds and 19,797 human proteins. PrePCI relies on a proteome-wide database of structural models based on both traditional modeling techniques and the AlphaFold Protein Structure Database. Sequence- and structural similarity-based metrics are established between template proteins, T, in the Protein Data Bank that bind compounds, C, and query proteins in the model database, Q. When the metrics exceed threshold values, it is assumed that C also binds to Q with a likelihood ratio (LR) derived from machine learning. If the relationship is based on structural similarity, the LR is based on a scoring function that measures the extent to which C is compatible with the binding site of Q as described in the LT-scanner algorithm. For every predicted complex derived in this way, chemical similarity based on the Tanimoto coefficient identifies other small molecules that may bind to Q. An overall LR for the binding of C to Q is obtained from Naive Bayesian statistics. The PrePCI database can be queried by entering a UniProt ID or gene name for a protein to obtain a list of compounds predicted to bind to it along with associated LRs. Alternatively, entering an identifier for the compound outputs a list of proteins it is predicted to bind. Specific applications of the database to lead discovery, elucidation of drug mechanism of action, and biological function annotation are described.

SCAMPER: Accurate Type-Specific Prediction of Calcium-Binding Residues Using Sequence-Derived Features

Understanding molecular mechanisms involved in calcium-protein interactions and modeling corresponding docking rely on the accurate identification of calcium-binding residues (CaBRs). The defects of experimentally annotating protein functions enhances the development of computational approaches that correctly identify calcium-binding interactions. Studies have reported that current methods severely cross-predict residues that interact with other types of molecules (e.g., nucleic acids, proteins, and small ligands) as CaBRs. In this study, a novel predictor named SCAMPER (Selective CAlciuM-binding PrEdictoR) is proposed for the accurate and specific prediction of CaBRs. SCAMPER is designed using newly compiled dataset with complete UniProt sequences and annotations, which include calcium-binding, nucleic acid-binding, protein-binding, and small ligand-binding residues. We use a novel designed two-layer scheme to perform predictions as well as penalize cross-predictions. Empirical tests on an independent test dataset reveals that the proposed method significantly outperforms state-of-the-art predictors. SCAMPER is proved to be capable of distinguishing CaBRs from different types of metal-ion binding residues. We further perform CaBRs predictions on the whole human proteome, and use the results to hypothesize calcium-binding proteins (CaBPs). The latest experimental verified CaBPs and GO analysis prove the accuracy of our predictions. We implement the proposed method and share the data at http://www.inforstation.com/webservers/SCAMPER/.

RCSB Protein Data Bank (RCSB.org): delivery of experimentally-determined PDB structures alongside one million computed structure models of proteins from artificial intelligence/machine learning

The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB), founding member of the Worldwide Protein Data Bank (wwPDB), is the US data center for the open-access PDB archive. As wwPDB-designated Archive Keeper, RCSB PDB is also responsible for PDB data security. Annually, RCSB PDB serves >10 000 depositors of three-dimensional (3D) biostructures working on all permanently inhabited continents. RCSB PDB delivers data from its research-focused RCSB.org web portal to many millions of PDB data consumers based in virtually every United Nations-recognized country, territory, etc. This Database Issue contribution describes upgrades to the research-focused RCSB.org web portal that created a one-stop-shop for open access to ∼200 000 experimentally-determined PDB structures of biological macromolecules alongside >1 000 000 incorporated Computed Structure Models (CSMs) predicted using artificial intelligence/machine learning methods. RCSB.org is a 'living data resource.' Every PDB structure and CSM is integrated weekly with related functional annotations from external biodata resources, providing up-to-date information for the entire corpus of 3D biostructure data freely available from RCSB.org with no usage limitations. Within RCSB.org, PDB structures and the CSMs are clearly identified as to their provenance and reliability. Both are fully searchable, and can be analyzed and visualized using the full complement of RCSB.org web portal capabilities.

Pathogenicity Prediction of Single Amino Acid Variants With Machine Learning Model Based on Protein Structural Energies

The most popular tools for predicting pathogenicity of single amino acid variants (SAVs) were developed based on sequence-based techniques. SAVs may change protein structure and function. In the context of van der Waals force and disulfide bridge calculations, no method directly predicts the impact of mutations on the energies of the protein structure. Here, we combined machine learning methods and energy scores of protein structures calculated by Rosetta Energy Function 2015 to predict SAV pathogenicity. The accuracy level of our model (0.76) is higher than that of six prediction tools. Further analyses revealed that the differential reference energies, attractive energies, and solvation of polar atoms between wildtype and mutant side-chains played essential roles in distinguishing benign from pathogenic variants. These features indicated the physicochemical properties of amino acids, which were observed in 3D structures instead of sequences. We added 16 features to Rhapsody (the prediction tool we used for our data set) and consequently improved its performance. The results indicated that these energy scores were more appropriate and more detailed representations of the pathogenicity of SAVs.

Overview and applications of map and model validation tools in the CCP-EM software suite

Cryogenic electron microscopy (cryo-EM) has recently been established as a powerful technique for solving macromolecular structures. Although the best resolutions achievable are improving, a significant majority of data are still resolved at resolutions worse than 3 Å, where it is non-trivial to build or fit atomic models. The map reconstructions and atomic models derived from the maps are also prone to errors accumulated through the different stages of data processing. Here, we highlight the need to evaluate both model geometry and fit to data at different resolutions. Assessment of cryo-EM structures from SARS-CoV-2 highlights a bias towards optimising the model geometry to agree with the most common conformations, compared to the agreement with data. We present the CoVal web service which provides multiple validation metrics to reflect the quality of atomic models derived from cryo-EM data of structures from SARS-CoV-2. We demonstrate that further refinement can lead to improvement of the agreement with data without the loss of geometric quality. We also discuss the recent CCP-EM developments aimed at addressing some of the current shortcomings.

Enhancing Protein Function Prediction Performance by Utilizing AlphaFold-Predicted Protein Structures

The structure of a protein is of great importance in determining its functionality, and this characteristic can be leveraged to train data-driven prediction models. However, the limited number of available protein structures severely limits the performance of these models. AlphaFold2 and its open-source data set of predicted protein structures have provided a promising solution to this problem, and these predicted structures are expected to benefit the model performance by increasing the number of training samples. In this work, we constructed a new data set that acted as a benchmark and implemented a state-of-the-art structure-based approach for determining whether the performance of the function prediction model can be improved by putting additional AlphaFold-predicted structures into the training set and further compared the performance differences between two models separately trained with real structures only and AlphaFold-predicted structures only. Experimental results indicated that structure-based protein function prediction models could benefit from virtual training data consisting of AlphaFold-predicted structures. First, model performances were improved in all three categories of Gene Ontology terms (GO terms) after adding predicted structures as training samples. Second, the model trained only on AlphaFold-predicted virtual samples achieved comparable performances to the model based on experimentally solved real structures, suggesting that predicted structures were almost equally effective in predicting protein functionality.

TMbed: transmembrane proteins predicted through language model embeddings

BACKGROUND

Despite the immense importance of transmembrane proteins (TMP) for molecular biology and medicine, experimental 3D structures for TMPs remain about 4-5 times underrepresented compared to non-TMPs. Today's top methods such as AlphaFold2 accurately predict 3D structures for many TMPs, but annotating transmembrane regions remains a limiting step for proteome-wide predictions.

RESULTS

Here, we present TMbed, a novel method inputting embeddings from protein Language Models (pLMs, here ProtT5), to predict for each residue one of four classes: transmembrane helix (TMH), transmembrane strand (TMB), signal peptide, or other. TMbed completes predictions for entire proteomes within hours on a single consumer-grade desktop machine at performance levels similar or better than methods, which are using evolutionary information from multiple sequence alignments (MSAs) of protein families. On the per-protein level, TMbed correctly identified 94 ± 8% of the beta barrel TMPs (53 of 57) and 98 ± 1% of the alpha helical TMPs (557 of 571) in a non-redundant data set, at false positive rates well below 1% (erred on 30 of 5654 non-membrane proteins). On the per-segment level, TMbed correctly placed, on average, 9 of 10 transmembrane segments within five residues of the experimental observation. Our method can handle sequences of up to 4200 residues on standard graphics cards used in desktop PCs (e.g., NVIDIA GeForce RTX 3060).

CONCLUSIONS

Based on embeddings from pLMs and two novel filters (Gaussian and Viterbi), TMbed predicts alpha helical and beta barrel TMPs at least as accurately as any other method but at lower false positive rates. Given the few false positives and its outstanding speed, TMbed might be ideal to sieve through millions of 3D structures soon to be predicted, e.g., by AlphaFold2.

Global Analysis of Deep Learning Prediction Using Large-Scale In-House Kinome-Wide Profiling Data

In drug discovery, the prediction of activity and absorption, distribution, metabolism, excretion, and toxicity parameters is one of the most important approaches in determining which compound to synthesize next. In recent years, prediction methods based on deep learning as well as non-deep learning approaches have been established, and a number of applications to drug discovery have been reported by various companies and organizations. In this research, we performed activity prediction using deep learning and non-deep learning methods on in-house assay data for several hundred kinases and compared and discussed the prediction results. We found that the prediction accuracy of the single-task graph neural network (GNN) model was generally lower than that of the non-deep learning model (LightGBM), but the multitask GNN model, which combined data from other kinases, comprehensively outperformed LightGBM. In addition, the extrapolative validity of the multitask model was verified by using it for prediction on known kinase ligands. We observed an overlap between characteristic protein-ligand interaction sites and the atoms that are important for prediction. By building appropriate models based on the conditions of the data set and analyzing the feature importance of the prediction results, a ligand-based prediction method may be used not only for activity prediction but also for drug design.

HPMPdb: a machine learning-ready database of protein molecular phenotypes associated to human missense variants

Current human Single Amino acid Variants (SAVs) databases provide a link between a SAVs and their effect on the carrier individual phenotype, often dividing them into Deleterious/Neutral variants. This is a very coarse-grained description of the genotype-to-phenotype relationship because it relies on un-realistic assumptions such as the perfect Mendelian behavior of each SAV and considers only dichotomic phenotypes. Moreover, the link between the effect of a SAV on a protein (its molecular phenotype) and the individual phenotype is often very complex, because multiple level of biological abstraction connect the protein and individual level phenotypes. Here we present HPMPdb, a manually curated database containing human SAVs associated with the detailed description of the molecular phenotype they cause on the affected proteins. With particular regards to machine learning (ML), this database can be used to let researchers go beyond the existing Deleterious/Neutral prediction paradigm, allowing them to build molecular phenotype predictors instead. Our class labels describe in a succinct way the effects that each SAV has on 15 protein molecular phenotypes, such as protein-protein interaction, small molecules binding, function, post-translational modifications (PTMs), sub-cellular localization, mimetic PTM, folding and protein expression. Moreover, we provide researchers with all necessary means to re-producibly train and test their models on our database. The webserver and the data described in this paper are available at hpmp.esat.kuleuven.be.

•

Current variant-effect predictors perform a coarse-grained modeling and rely on unrealistic assumptions.

•

The link between the effect of a variant and the individual phenotype is complex. It would be more intuitive to predict the molecular phenotype that each variant causes on the carrier protein.

•

HPMP is a manually curated database containing human variants associated with the molecular phenotype they cause on the affected proteins.

•

We manually translated variants from Uniprot into 15 Machine Learning-ready labels describing the affected protein molecular phenotype.

•

The goal of HPMP is to allow researchers to go beyond the existing variant-effect prediction paradigm and allow them to build molecular phenotype predictors instead.

•

The webserver and the data described in this paper are available at hpmp.esat.kuleuven.be

PSnpBind: a database of mutated binding site protein-ligand complexes constructed using a multithreaded virtual screening workflow

A key concept in drug design is how natural variants, especially the ones occurring in the binding site of drug targets, affect the inter-individual drug response and efficacy by altering binding affinity. These effects have been studied on very limited and small datasets while, ideally, a large dataset of binding affinity changes due to binding site single-nucleotide polymorphisms (SNPs) is needed for evaluation. However, to the best of our knowledge, such a dataset does not exist. Thus, a reference dataset of ligands binding affinities to proteins with all their reported binding sites’ variants was constructed using a molecular docking approach. Having a large database of protein–ligand complexes covering a wide range of binding pocket mutations and a large small molecules’ landscape is of great importance for several types of studies. For example, developing machine learning algorithms to predict protein–ligand affinity or a SNP effect on it requires an extensive amount of data. In this work, we present PSnpBind: A large database of 0.6 million mutated binding site protein–ligand complexes constructed using a multithreaded virtual screening workflow. It provides a web interface to explore and visualize the protein–ligand complexes and a REST API to programmatically access the different aspects of the database contents. PSnpBind is open source and freely available at https://psnpbind.org.

PIPENN: protein interface prediction from sequence with an ensemble of neural nets

MOTIVATION

The interactions between proteins and other molecules are essential to many biological and cellular processes. Experimental identification of interface residues is a time-consuming, costly and challenging task, while protein sequence data are ubiquitous. Consequently, many computational and machine learning approaches have been developed over the years to predict such interface residues from sequence. However, the effectiveness of different Deep Learning (DL) architectures and learning strategies for protein-protein, protein-nucleotide and protein-small molecule interface prediction has not yet been investigated in great detail. Therefore, we here explore the prediction of protein interface residues using six DL architectures and various learning strategies with sequence-derived input features.

RESULTS

We constructed a large dataset dubbed BioDL, comprising protein-protein interactions from the PDB, and DNA/RNA and small molecule interactions from the BioLip database. We also constructed six DL architectures, and evaluated them on the BioDL benchmarks. This shows that no single architecture performs best on all instances. An ensemble architecture, which combines all six architectures, does consistently achieve peak prediction accuracy. We confirmed these results on the published benchmark set by Zhang and Kurgan (ZK448), and on our own existing curated homo- and heteromeric protein interaction dataset. Our PIPENN sequence-based ensemble predictor outperforms current state-of-the-art sequence-based protein interface predictors on ZK448 on all interaction types, achieving an AUC-ROC of 0.718 for protein-protein, 0.823 for protein-nucleotide and 0.842 for protein-small molecule.

AVAILABILITY AND IMPLEMENTATION

Source code and datasets are available at https://github.com/ibivu/pipenn/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

PPI-HotspotDB: Database of Protein-Protein Interaction Hot Spots

Single-point mutations of certain residues (so-called hot spots) impair/disrupt protein-protein interactions (PPIs), leading to pathogenesis and drug resistance. Conventionally, a PPI-hot spot is identified when its replacement decreased the binding free energy significantly, generally by ≥2 kcal/mol. The relatively few mutations with such a significant binding free energy drop limited the number of distinct PPI-hot spots. By defining PPI-hot spots based on mutations that have been manually curated in UniProtKB to significantly impair/disrupt PPIs in addition to binding free energy changes, we have greatly expanded the number of distinct PPI-hot spots by an order of magnitude. These experimentally determined PPI-hot spots along with available structures have been collected in a database called PPI-Hotspot^DB. We have applied the PPI-Hotspot^DB to create a nonredundant benchmark, PPI-Hotspot+PDB^BM, for assessing methods to predict PPI-hot spots using the free structure as input. PPI-Hotspot^DB will benefit the design of mutagenesis experiments and development of PPI-hot spot prediction methods. The database and benchmark are freely available at https://ppihotspot.limlab.dnsalias.org.

Kincore: a web resource for structural classification of protein kinases and their inhibitors

The active form of kinases is shared across different family members, as are several commonly observed inactive forms. We previously performed a clustering of the conformation of the activation loop of all protein kinase structures in the Protein Data Bank (PDB) into eight classes based on the dihedral angles that place the Phe side chain of the DFG motif at the N-terminus of the activation loop. Our clusters are strongly associated with the placement of the activation loop, the C-helix, and other structural elements of kinases. We present Kincore, a web resource providing access to our conformational assignments for kinase structures in the PDB. While other available databases provide conformational states or drug type but not both, KinCore includes the conformational state and the inhibitor type (Type 1, 1.5, 2, 3, allosteric) for each kinase chain. The user can query and browse the database using these attributes or determine the conformational labels of a kinase structure using the web server or a standalone program. The database and labeled structure files can be downloaded from the server. Kincore will help in understanding the conformational dynamics of these proteins and guide development of inhibitors targeting specific states. Kincore is available at http://dunbrack.fccc.edu/kincore.

Profiles of Natural and Designed Protein-Like Sequences Effectively Bridge Protein Sequence Gaps: Implications in Distant Homology Detection

Sequence-based approaches are fundamental to guide experimental investigations in obtaining structural and/or functional insights into uncharacterized protein families. Powerful profile-based sequence search methods rely on a sequence space continuum to identify non-trivial relationships through homology detection. The computational design of protein-like sequences that serve as "artificial linkers" is useful in identifying relationships between distant members of a structural fold. Such sequences act as intermediates and guide homology searches between distantly related proteins. Here, we describe an approach that represents natural intermediate sequences and designed protein-like sequences as HMM (Hidden Markov Models) profiles, to improve the sensitivity of existing search methods. Searches made within the "Profile database" were shown to recognize the parent structural fold for 90% of the search queries at query coverage better than 60%. For 1040 protein families with no available structure, fold associations were made through searches in the database of natural and designed sequence profiles. Most of the associations were made with the Alpha-alpha superhelix, Transmembrane beta-barrels, TIM barrel, and Immunoglobulin-like beta-sandwich folds. For 11 domain families of unknown functions, we provide confident fold associations using the profiles of designed sequences and a consensus from other fold recognition methods. For two DUFs (Domain families of Unknown Functions), we performed detailed functional annotation through comparisons with characterized templates of families of known function.

Yeast metabolic innovations emerged via expanded metabolic network and gene positive selection

Yeasts are known to have versatile metabolic traits, while how these metabolic traits have evolved has not been elucidated systematically. We performed integrative evolution analysis to investigate how genomic evolution determines trait generation by reconstructing genome-scale metabolic models (GEMs) for 332 yeasts. These GEMs could comprehensively characterize trait diversity and predict enzyme functionality, thereby signifying that sequence-level evolution has shaped reaction networks towards new metabolic functions. Strikingly, using GEMs, we can mechanistically map different evolutionary events, e.g. horizontal gene transfer and gene duplication, onto relevant subpathways to explain metabolic plasticity. This demonstrates that gene family expansion and enzyme promiscuity are prominent mechanisms for metabolic trait gains, while GEM simulations reveal that additional factors, such as gene loss from distant pathways, contribute to trait losses. Furthermore, our analysis could pinpoint to specific genes and pathways that have been under positive selection and relevant for the formulation of complex metabolic traits, i.e. thermotolerance and the Crabtree effect. Our findings illustrate how multidimensional evolution in both metabolic network structure and individual enzymes drives phenotypic variations.

Structure-based prediction of post-translational modification cross-talk within proteins using complementary residue- and residue pair-based features

Post-translational modification (PTM)-based regulation can be mediated not only by the modification of a single residue but also by the interplay of different modifications. Accurate prediction of PTM cross-talk is a highly challenging issue and is in its infant stage. Especially, less attention has been paid to the structural preferences (except intrinsic disorder and spatial proximity) of cross-talk pairs and the characteristics of individual residues involved in cross-talk, which may restrict the improvement of the prediction accuracy. Here we report a structure-based algorithm called PCTpred to improve the PTM cross-talk prediction. The comprehensive residue- and residue pair-based features were designed for paired PTM sites at the sequence and structural levels. Through feature selection, we reserved 23 newly introduced descriptors and 3 traditional descriptors to develop a sequence-based predictor PCTseq and a structure-based predictor PCTstr, both of which were integrated to construct our final prediction model. According to pair- and protein-based evaluations, PCTpred yielded area under the curve values of approximately 0.9 and 0.8, respectively. Even when removing the distance preference of samples or using the input of modeled structures, our prediction performance was maintained or moderately reduced. PCTpred displayed stable and reliable improvements over the state-of-the-art methods based on various evaluations. The source code and data set are freely available at https://github.com/Liulab-HZAU/PCTpred or http://liulab.hzau.edu.cn/PCTpred/.

Ankyrin repeats in context with human population variation

Ankyrin protein repeats bind to a wide range of substrates and are one of the most common protein motifs in nature. Here, we collate a high-quality alignment of 7,407 ankyrin repeats and examine for the first time, the distribution of human population variants from large-scale sequencing of healthy individuals across this family. Population variants are not randomly distributed across the genome but are constrained by gene essentiality and function. Accordingly, we interpret the population variants in context with evolutionary constraint and structural features including secondary structure, accessibility and protein-protein interactions across 383 three-dimensional structures of ankyrin repeats. We find five positions that are highly conserved across homologues and also depleted in missense variants within the human population. These positions are significantly enriched in intra-domain contacts and so likely to be key for repeat packing. In contrast, a group of evolutionarily divergent positions are found to be depleted in missense variants in human and significantly enriched in protein-protein interactions. Our analysis also suggests the domain has three, not two surfaces, each with different patterns of enrichment in protein-substrate interactions and missense variants. Our findings will be of interest to those studying or engineering ankyrin-repeat containing proteins as well as those interpreting the significance of disease variants.

D3DistalMutation: a Database to Explore the Effect of Distal Mutations on Enzyme Activity

Enzyme activity is affected by amino acid mutations, particularly mutations near the active site. Increasing evidence has shown that distal mutations more than 10 Å away from the active site may significantly affect enzyme activity. However, it is difficult to study the enzyme regulation mechanism of distal mutations due to the lack of a systematic collection of three-dimensional (3D) structures, highlighting distal mutation site and the corresponding enzyme activity change. Therefore, we constructed a distal mutation database, namely, D3DistalMutation, which relates the distal mutation to enzyme activity. As a result, we observed that approximately 80% of distal mutations could affect enzyme activity and 72.7% of distal mutations would decrease or abolish enzyme activity in D3DistalMutation. Only 6.6% of distal mutations in D3DistalMutation could increase enzyme activity, which have great potential to the industrial field. Among these mutations, the Y to F, S to D, and T to D mutations are most likely to increase enzyme activity, which sheds some light on industrial catalysis. Distal mutations decreasing enzyme activity in the allosteric pocket play an indispensable role in allosteric drug design. In addition, the pockets in the enzyme structures are provided to explore the enzyme regulation mechanism of distal mutations. D3DistalMutation is accessible free of charge at https://www.d3pharma.com/D3DistalMutation/index.php.

Immuno-informatics design of a multimeric epitope peptide based vaccine targeting SARS-CoV-2 spike glycoprotein

Developing an efficacious vaccine for SARS-CoV-2 infection is critical to stemming COVID-19 fatalities and providing the global community with immune protection. We have used a bioinformatic approach to aid in designing an epitope peptide-based vaccine against the spike protein of the virus. Five antigenic B cell epitopes with viable antigenicity and a total of 27 discontinuous B cell epitopes were mapped out structurally in the spike protein for antibody recognition. We identified eight CD8+ T cell 9-mers and 12 CD4+ T cell 14-15-mer as promising candidate epitopes putatively restricted by a large number of MHC I and II alleles, respectively. We used this information to construct an in silico chimeric peptide vaccine whose translational rate was highly expressed when cloned in pET28a (+) vector. With our In silico test, the vaccine construct was predicted to elicit high antigenicity and cell-mediated immunity when given as a homologous prime-boost, triggering of toll-like receptor 5 by the adjuvant linker. The vaccine was also characterized by an increase in IgM and IgG and an array of Th1 and Th2 cytokines. Upon in silico challenge with SARS-CoV-2, there was a decrease in antigen levels using our immune simulations. We, therefore, propose that potential vaccine designs consider this approach.

Large-scale discovery of protein interactions at residue resolution using co-evolution calculated from genomic sequences

Increasing numbers of protein interactions have been identified in high-throughput experiments, but only a small proportion have solved structures. Recently, sequence coevolution-based approaches have led to a breakthrough in predicting monomer protein structures and protein interaction interfaces. Here, we address the challenges of large-scale interaction prediction at residue resolution with a fast alignment concatenation method and a probabilistic score for the interaction of residues. Importantly, this method (EVcomplex2) is able to assess the likelihood of a protein interaction, as we show here applied to large-scale experimental datasets where the pairwise interactions are unknown. We predict 504 interactions de novo in the E. coli membrane proteome, including 243 that are newly discovered. While EVcomplex2 does not require available structures, coevolving residue pairs can be used to produce structural models of protein interactions, as done here for membrane complexes including the Flagellar Hook-Filament Junction and the Tol/Pal complex.

Our understanding of the residue-level details of protein interactions remains incomplete. Here, the authors show sequence coevolution can be used to infer interacting proteins with residue-level details, including predicting 467 interactions de novo in the Escherichia coli cell envelope proteome.

FLIPPER: Predicting and Characterizing Linear Interacting Peptides in the Protein Data Bank

A large fraction of peptides or protein regions are disordered in isolation and fold upon binding. These regions, also called MoRFs, SLiMs or LIPs, are often associated with signaling and regulation processes. However, despite their importance, only a limited number of examples are available in public databases and their automatic detection at the proteome level is problematic. Here we present FLIPPER, an automatic method for the detection of structurally linear sub-regions or peptides that interact with another chain in a protein complex. FLIPPER is a random forest classification that takes the protein structure as input and provides the propensity of each amino acid to be part of a LIP region. Models are built taking into consideration structural features such as intra- and inter-chain contacts, secondary structure, solvent accessibility in both bound and unbound state, structural linearity and chain length. FLIPPER is accurate when evaluated on non-redundant independent datasets, 99% precision and 99% sensitivity on PixelDB-25 and 87% precision and 88% sensitivity on DIBS-25. Finally, we used FLIPPER to process the entire Protein Data Bank and identified different classes of LIPs based on different binding modes and partner molecules. We provide a detailed description of these LIP categories and show that a large fraction of these regions are not detected by disorder predictors. All FLIPPER predictions are integrated in the MobiDB 4.0 database.

Protein dynamics analysis identifies candidate cancer driver genes and mutations in TCGA data

Recently, it has been showed that cancer missense mutations selectively target the neighborhood of hinge residues, which are key sites in protein dynamics. Here, we show that this approach can be extended to find previously unknown candidate mutations and genes. To this aim, we developed a computational pipeline to detect significantly enriched three-dimensional (3D) clustering of missense mutations around hinge residues. The hinge residues were detected by applying a Gaussian network model. By systematically analyzing the PanCancer compendium of somatic mutations in nearly 10 000 tumors from the Cancer Genome Atlas, we identified candidate genes and mutations in addition to well known ones. For instance, we found significantly enriched 3D clustering of missense mutations in known cancer genes including CDK4, CDKN2A, TCL1A, and MAPK1. Beside these known genes, we also identified significantly enriched 3D clustering of missense mutations around hinge residues in PLA2G4A, which may lead to excessive phosphorylation of the extracellular signal-regulated kinases. Furthermore, we demonstrated that hinge-based features improves pathogenicity prediction for missense mutations. Our results show that the consideration of clustering around hinge residues can help us explain the functional role of the mutations in known cancer genes and identify candidate genes.

GPCRdb in 2021: integrating GPCR sequence, structure and function

G protein-coupled receptors (GPCRs) form both the largest family of membrane proteins and drug targets, mediating the action of one-third of medicines. The GPCR database, GPCRdb serves >4 000 researchers every month and offers reference data, analysis of own or literature data, experiment design and dissemination of published datasets. Here, we describe new and updated GPCRdb resources with a particular focus on integration of sequence, structure and function. GPCRdb contains all human non-olfactory GPCRs (and >27 000 orthologs), G-proteins and arrestins. It includes over 2 000 drug and in-trial agents and nearly 200 000 ligands with activity and availability data. GPCRdb annotates all published GPCR structures (updated monthly), which are also offered in a refined version (with re-modeled missing/distorted regions and reverted mutations) and provides structure models of all human non-olfactory receptors in inactive, intermediate and active states. Mutagenesis data in the GPCRdb spans natural genetic variants, GPCR-G protein interfaces, ligand sites and thermostabilising mutations. A new sequence signature tool for identification of functional residue determinants has been added and two data driven tools to design ligand site mutations and constructs for structure determination have been updated extending their coverage of receptors and modifications. The GPCRdb is available at https://gpcrdb.org.

Sequence and Structure-Based Analysis of Specificity Determinants in Eukaryotic Protein Kinases

Protein kinases lie at the heart of cell-signaling processes and are often mutated in disease. Kinase target recognition at the active site is in part determined by a few amino acids around the phosphoacceptor residue. However, relatively little is known about how most preferences are encoded in the kinase sequence or how these preferences evolved. Here, we used alignment-based approaches to predict 30 specificity-determining residues (SDRs) for 16 preferences. These were studied with structural models and were validated by activity assays of mutant kinases. Cancer mutation data revealed that kinase SDRs are mutated more frequently than catalytic residues. We have observed that, throughout evolution, kinase specificity has been strongly conserved across orthologs but can diverge after gene duplication, as illustrated by the G protein-coupled receptor kinase family. The identified SDRs can be used to predict kinase specificity from sequence and aid in the interpretation of evolutionary or disease-related genomic variants.

Comprehensive characterization of amino acid positions in protein structures reveals molecular effect of missense variants

Interpretation of the colossal number of genetic variants identified from sequencing applications is one of the major bottlenecks in clinical genetics, with the inference of the effect of amino acid-substituting missense variations on protein structure and function being especially challenging. Here we characterize the three-dimensional (3D) amino acid positions affected in pathogenic and population variants from 1,330 disease-associated genes using over 14,000 experimentally solved human protein structures. By measuring the statistical burden of variations (i.e., point mutations) from all genes on 40 3D protein features, accounting for the structural, chemical, and functional context of the variations' positions, we identify features that are generally associated with pathogenic and population missense variants. We then perform the same amino acid-level analysis individually for 24 protein functional classes, which reveals unique characteristics of the positions of the altered amino acids: We observe up to 46% divergence of the class-specific features from the general characteristics obtained by the analysis on all genes, which is consistent with the structural diversity of essential regions across different protein classes. We demonstrate that the function-specific 3D features of the variants match the readouts of mutagenesis experiments for BRCA1 and PTEN, and positively correlate with an independent set of clinically interpreted pathogenic and benign missense variants. Finally, we make our results available through a web server to foster accessibility and downstream research. Our findings represent a crucial step toward translational genetics, from highlighting the impact of mutations on protein structure to rationalizing the variants' pathogenicity in terms of the perturbed molecular mechanisms.

Protein-protein and protein-nucleic acid binding residues important for common and rare sequence variants in human

BACKGROUND

Any two unrelated people differ by about 20,000 missense mutations (also referred to as SAVs: Single Amino acid Variants or missense SNV). Many SAVs have been predicted to strongly affect molecular protein function. Common SAVs (> 5% of population) were predicted to have, on average, more effect on molecular protein function than rare SAVs (< 1% of population). We hypothesized that the prevalence of effect in common over rare SAVs might partially be caused by common SAVs more often occurring at interfaces of proteins with other proteins, DNA, or RNA, thereby creating subgroup-specific phenotypes. We analyzed SAVs from 60,706 people through the lens of two prediction methods, one (SNAP2) predicting the effects of SAVs on molecular protein function, the other (ProNA2020) predicting residues in DNA-, RNA- and protein-binding interfaces.

RESULTS

Three results stood out. Firstly, SAVs predicted to occur at binding interfaces were predicted to more likely affect molecular function than those predicted as not binding (p value < 2.2 × 10^-16). Secondly, for SAVs predicted to occur at binding interfaces, common SAVs were predicted more strongly with effect on protein function than rare SAVs (p value < 2.2 × 10^-16). Restriction to SAVs with experimental annotations confirmed all results, although the resulting subsets were too small to establish statistical significance for any result. Thirdly, the fraction of SAVs predicted at binding interfaces differed significantly between tissues, e.g. urinary bladder tissue was found abundant in SAVs predicted at protein-binding interfaces, and reproductive tissues (ovary, testis, vagina, seminal vesicle and endometrium) in SAVs predicted at DNA-binding interfaces.

CONCLUSIONS

Overall, the results suggested that residues at protein-, DNA-, and RNA-binding interfaces contributed toward predicting that common SAVs more likely affect molecular function than rare SAVs.

A Sweep of Earth's Virome Reveals Host-Guided Viral Protein Structural Mimicry and Points to Determinants of Human Disease

Viruses deploy genetically encoded strategies to coopt host machinery and support viral replicative cycles. Here, we use protein structure similarity to scan for molecular mimicry, manifested by structural similarity between viral and endogenous host proteins, across thousands of cataloged viruses and hosts spanning broad ecological niches and taxonomic range, including bacteria, plants and fungi, invertebrates, and vertebrates. This survey identified over 6,000,000 instances of structural mimicry; more than 70% of viral mimics cannot be discerned through protein sequence alone. We demonstrate that the manner and degree to which viruses exploit molecular mimicry varies by genome size and nucleic acid type and identify 158 human proteins that are mimicked by coronaviruses, providing clues about cellular processes driving pathogenesis. Our observations point to molecular mimicry as a pervasive strategy employed by viruses and indicate that the protein structure space used by a given virus is dictated by the host proteome. A record of this paper's transparent peer review process is included in the Supplemental Information.

MISCAST: MIssense variant to protein StruCture Analysis web SuiTe

Human genome sequencing efforts have greatly expanded, and a plethora of missense variants identified both in patients and in the general population is now publicly accessible. Interpretation of the molecular-level effect of missense variants, however, remains challenging and requires a particular investigation of amino acid substitutions in the context of protein structure and function. Answers to questions like 'Is a variant perturbing a site involved in key macromolecular interactions and/or cellular signaling?', or 'Is a variant changing an amino acid located at the protein core or part of a cluster of known pathogenic mutations in 3D?' are crucial. Motivated by these needs, we developed MISCAST (missense variant to protein structure analysis web suite; http://miscast.broadinstitute.org/). MISCAST is an interactive and user-friendly web server to visualize and analyze missense variants in protein sequence and structure space. Additionally, a comprehensive set of protein structural and functional features have been aggregated in MISCAST from multiple databases, and displayed on structures alongside the variants to provide users with the biological context of the variant location in an integrated platform. We further made the annotated data and protein structures readily downloadable from MISCAST to foster advanced offline analysis of missense variants by a wide biological community.

Identifying molecular recognition features in intrinsically disordered regions of proteins by transfer learning

MOTIVATION

Protein intrinsic disorder describes the tendency of sequence residues to not fold into a rigid three-dimensional shape by themselves. However, some of these disordered regions can transition from disorder to order when interacting with another molecule in segments known as molecular recognition features (MoRFs). Previous analysis has shown that these MoRF regions are indirectly encoded within the prediction of residue disorder as low-confidence predictions [i.e. in a semi-disordered state P(D)≈0.5]. Thus, what has been learned for disorder prediction may be transferable to MoRF prediction. Transferring the internal characterization of protein disorder for the prediction of MoRF residues would allow us to take advantage of the large training set available for disorder prediction, enabling the training of larger analytical models than is currently feasible on the small number of currently available annotated MoRF proteins. In this paper, we propose a new method for MoRF prediction by transfer learning from the SPOT-Disorder2 ensemble models built for disorder prediction.

RESULTS

We confirm that directly training on the MoRF set with a randomly initialized model yields substantially poorer performance on independent test sets than by using the transfer-learning-based method SPOT-MoRF, for both deep and simple networks. Its comparison to current state-of-the-art techniques reveals its superior performance in identifying MoRF binding regions in proteins across two independent testing sets, including our new dataset of >800 protein chains. These test chains share <30% sequence similarity to all training and validation proteins used in SPOT-Disorder2 and SPOT-MoRF, and provide a much-needed large-scale update on the performance of current MoRF predictors. The method is expected to be useful in locating functional disordered regions in proteins.

AVAILABILITY AND IMPLEMENTATION

SPOT-MoRF and its data are available as a web server and as a standalone program at: http://sparks-lab.org/jack/server/SPOT-MoRF/index.php.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

PSCAN: Spatial scan tests guided by protein structures improve complex disease gene discovery and signal variant detection

Germline disease-causing variants are generally more spatially clustered in protein 3-dimensional structures than benign variants. Motivated by this tendency, we develop a fast and powerful protein-structure-based scan (PSCAN) approach for evaluating gene-level associations with complex disease and detecting signal variants. We validate PSCAN's performance on synthetic data and two real data sets for lipid traits and Alzheimer's disease. Our results demonstrate that PSCAN performs competitively with existing gene-level tests while increasing power and identifying more specific signal variant sets. Furthermore, PSCAN enables generation of hypotheses about the molecular basis for the associations in the context of protein structures and functional domains.

An in silico approach to identification, categorization and prediction of nucleic acid binding proteins

The interaction between proteins and nucleic acid plays an important role in many processes, such as transcription, translation and DNA repair. The mechanisms of related biological events can be understood by exploring the function of proteins in these interactions. The number of known protein sequences has increased rapidly in recent years, but the databases for describing the structure and function of protein have unfortunately grown quite slowly. Thus, improving such databases is meaningful for predicting protein-nucleic acid interactions. Furthermore, the mechanism of related biological events, such as viral infection or designing novel drug targets, can be further understood by understanding the function of proteins in these interactions. The information for each sequence, including its function and interaction sites, were collected and identified, and a database called PNIDB was built. The proteins in PNIDB were grouped into 27 classes, such as transcription, immune system, and structural protein, etc. The function of each protein was then predicted using a machine learning method. Using our method, the predictor was trained on labeled sequences, and then the function of a protein was predicted based on the trained classifier. The prediction accuracy achieved a score of 77.43% by 10-fold cross validation.

Identification of Intrinsic Disorder in Complexes from the Protein Data Bank

Background: Intrinsically disordered proteins or regions (IDPs or IDRs) lack stable structures in solution, yet often fold upon binding with partners. IDPs or IDRs are highly abundant in all proteomes and represent a significant modification of sequence → structure → function paradigm. The Protein Data Bank (PDB) includes complexes containing disordered segments bound to globular proteins, but the molecular mechanisms of such binding interactions remain largely unknown. Results: In this study, we present the results of various disorder predictions on a nonredundant set of PDB complexes. In contrast to their structural appearances, many PDB proteins were predicted to be disordered when separated from their binding partners. These predicted-to-be-disordered proteins were observed to form structures depending upon various factors, including heterogroup binding, protein/DNA/RNA binding, disulfide bonds, and ion binding. Conclusions: This study collects many examples of disorder-to-order transition in IDP complex formation, thus revealing the unusual structure–function relationships of IDPs and providing an additional support for the newly proposed paradigm of the sequence → IDP/IDR ensemble → function.

Prediction of DNA-Binding Residues in Local Segments of Protein Sequences with Fuzzy Cognitive Maps

While protein-DNA interactions are crucial for a wide range of cellular functions, only a small fraction of these interactions was annotated to date. One solution to close this annotation gap is to employ computational methods that accurately predict protein-DNA interactions from widely available protein sequences. We present and empirically test first-of-its-kind predictor of DNA-binding residues in local segments of protein sequences that relies on the Fuzzy Cognitive Map (FCM) model. The FCM model uses information about putative solvent accessibility, evolutionary conservation, and relative propensities of amino acid to interact with DNA to generate putative DNA-binding residues. Empirical tests on a benchmark dataset reveal that the FCM model secures AUC = 0.72 and outperforms recently released hybridNAP predictor and several popular machine learning methods including Support Vector Machines, Naïve Bayes, and k-Nearest Neighbor. The improvements in the predictive performance result from an intrinsic feature of FCMs that incorporate relations between the input features, besides the relations between the inputs and output that are modelled by other algorithms. We also empirically demonstrate that use of a short sliding window results in further improvements in the predictive quality. The funDNApred webserver that implements the FCM predictor is available at http://biomine.cs.vcu.edu/servers/funDNApred/.