SCOPe: classification of large macromolecular structures in the structural classification of proteins-extended database

Sensitive remote homology search by local alignment of small positional embeddings from protein language models

Accurately detecting distant evolutionary relationships between proteins remains an ongoing challenge in bioinformatics. Search methods based on primary sequence struggle to accurately detect homology between sequences with less than 20% amino acid identity. Profile- and structure-based strategies extend sensitive search capabilities into this twilight zone of sequence similarity but require slow pre-processing steps. Recently, whole-protein and positional embeddings from deep neural networks have shown promise for providing sensitive sequence comparison and annotation at long evolutionary distances. Embeddings are generally faster to compute than profiles and predicted structures but still suffer several drawbacks related to the ability of whole-protein embeddings to discriminate domain-level homology, and the database size and search speed of methods using positional embeddings. In this work, we show that low-dimensionality positional embeddings can be used directly in speed-optimized local search algorithms. As a proof of concept, we use the ESM2 3B model to convert primary sequences directly into the 3D interaction (3Di) alphabet or amino acid profiles and use these embeddings as input to the highly optimized Foldseek, HMMER3, and HH-suite search algorithms. Our results suggest that positional embeddings as small as a single byte can provide sufficient information for dramatically improved sensitivity over amino acid sequence searches without sacrificing search speed.

Fast and accurate protein structure search with Foldseek

As structure prediction methods are generating millions of publicly available protein structures, searching these databases is becoming a bottleneck. Foldseek aligns the structure of a query protein against a database by describing tertiary amino acid interactions within proteins as sequences over a structural alphabet. Foldseek decreases computation times by four to five orders of magnitude with 86%, 88% and 133% of the sensitivities of Dali, TM-align and CE, respectively.

Origins and Functional Significance of Eukaryotic Protein Folds

Folds are the architecture and topology of a protein domain. Categories of folds are very few compared to the astronomical number of sequences. Eukaryotes have more protein folds than Archaea and Bacteria. These folds are of two types: shared with Archaea and/or Bacteria on one hand and specific to eukaryotic clades on the other hand. The first kind of folds is inherited from the first endosymbiosis and confirms the mixed origin of eukaryotes. In a dataset of 1073 folds whose presence or absence has been evidenced among 210 species equally distributed in the three super-kingdoms, we have identified 28 eukaryotic folds unambiguously inherited from Bacteria and 40 eukaryotic folds unambiguously inherited from Archaea. Compared to previous studies, the repartition of informational function is higher than expected for folds originated from Bacteria and as high as expected for folds inherited from Archaea. The second type of folds is specifically eukaryotic and associated with an increase of new folds within eukaryotes distributed in particular clades. Reconstructed ancestral states coupled with dating of each node on the tree of life provided fold appearance rates. The rate is on average twice higher within Eukaryota than within Bacteria or Archaea. The highest rates are found in the origins of eukaryotes, holozoans, metazoans, metazoans stricto sensu, and vertebrates: the roots of these clades correspond to bursts of fold evolution. We could correlate the functions of some of the fold synapomorphies within eukaryotes with significant evolutionary events. Among them, we find evidence for the rise of multicellularity, adaptive immune system, or virus folds which could be linked to an ecological shift made by tetrapods.

Retracing the evolution of a modern periplasmic binding protein

Investigating the evolution of structural features in modern multidomain proteins helps to understand their immense diversity and functional versatility. The class of periplasmic binding proteins (PBPs) offers an opportunity to interrogate one of the main processes driving diversification: the duplication and fusion of protein sequences to generate new architectures. The symmetry of their two-lobed topology, their mechanism of binding, and the organization of their operon structure led to the hypothesis that PBPs arose through a duplication and fusion event of a single common ancestor. To investigate this claim, we set out to reverse the evolutionary process and recreate the structural equivalent of a single-lobed progenitor using ribose-binding protein (RBP) as our model. We found that this modern PBP can be deconstructed into its lobes, producing two proteins that represent possible progenitor halves. The isolated halves of RBP are well folded and monomeric proteins, albeit with a lower thermostability, and do not retain the original binding function. However, the two entities readily form a heterodimer in vitro and in-cell. The x-ray structure of the heterodimer closely resembles the parental protein. Moreover, the binding function is fully regained upon formation of the heterodimer with a ligand affinity similar to that observed in the modern RBP. This highlights how a duplication event could have given rise to a stable and functional PBP-like fold and provides insights into how more complex functional structures can evolve from simpler molecular components.

PreHom-PCLM: protein remote homology detection by combing motifs and protein cubic language model

Protein remote homology detection is essential for structure prediction, function prediction, disease mechanism understanding, etc. The remote homology relationship depends on multiple protein properties, such as structural information and local sequence patterns. Previous studies have shown the challenges for predicting remote homology relationship by protein features at sequence level (e.g. position-specific score matrix). Protein motifs have been used in structure and function analysis due to their unique sequence patterns and implied structural information. Therefore, designing a usable architecture to fuse multiple protein properties based on motifs is urgently needed to improve protein remote homology detection performance. To make full use of the characteristics of motifs, we employed the language model called the protein cubic language model (PCLM). It combines multiple properties by constructing a motif-based neural network. Based on the PCLM, we proposed a predictor called PreHom-PCLM by extracting and fusing multiple motif features for protein remote homology detection. PreHom-PCLM outperforms the other state-of-the-art methods on the test set and independent test set. Experimental results further prove the effectiveness of multiple features fused by PreHom-PCLM for remote homology detection. Furthermore, the protein features derived from the PreHom-PCLM show strong discriminative power for proteins from different structural classes in the high-dimensional space. Availability and Implementation: http://bliulab.net/PreHom-PCLM.

Investigating the conformational landscape of AlphaFold2-predicted protein kinase structures

Summary

Protein kinases are a family of signaling proteins, crucial for maintaining cellular homeostasis. When dysregulated, kinases drive the pathogenesis of several diseases, and are thus one of the largest target categories for drug discovery. Kinase activity is tightly controlled by switching through several active and inactive conformations in their catalytic domain. Kinase inhibitors have been designed to engage kinases in specific conformational states, where each conformation presents a unique physico-chemical environment for therapeutic intervention. Thus, modeling kinases across conformations can enable the design of novel and optimally selective kinase drugs. Due to the recent success of AlphaFold2 in accurately predicting the 3D structure of proteins based on sequence, we investigated the conformational landscape of protein kinases as modeled by AlphaFold2. We observed that AlphaFold2 is able to model several kinase conformations across the kinome, however, certain conformations are only observed in specific kinase families. Furthermore, we show that the per residue predicted local distance difference test can capture information describing structural flexibility of kinases. Finally, we evaluated the docking performance of AlphaFold2 kinase structures for enriching known ligands. Taken together, we see an opportunity to leverage AlphaFold2 models for structure-based drug discovery against kinases across several pharmacologically relevant conformational states.

Availability and implementation

All code used in the analysis is freely available at https://github.com/Harmonic-Discovery/AF2-kinase-conformational-landscape.

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.

Protein diversification through post-translational modifications, alternative splicing, and gene duplication

Proteins provide the basis for cellular function. Having multiple versions of the same protein within a single organism provides a way of regulating its activity or developing novel functions. Post-translational modifications of proteins, by means of adding/removing chemical groups to amino acids, allow for a well-regulated and controlled way of generating functionally distinct protein species. Alternative splicing is another method with which organisms possibly generate new isoforms. Additionally, gene duplication events throughout evolution generate multiple paralogs of the same genes, resulting in multiple versions of the same protein within an organism. In this review, we discuss recent advancements in the study of these three methods of protein diversification and provide illustrative examples of how they affect protein structure and function.

Plant-on-chip: Core morphogenesis processes in the tiny plant Wolffia australiana

A plant can be thought of as a colony comprising numerous growth buds, each developing to its own rhythm. Such lack of synchrony impedes efforts to describe core principles of plant morphogenesis, dissect the underlying mechanisms, and identify regulators. Here, we use the minimalist known angiosperm to overcome this challenge and provide a model system for plant morphogenesis. We present a detailed morphological description of the monocot Wolffia australiana, as well as high-quality genome information. Further, we developed the plant-on-chip culture system and demonstrate the application of advanced technologies such as single-nucleus RNA-sequencing, protein structure prediction, and gene editing. We provide proof-of-concept examples that illustrate how W. australiana can decipher the core regulatory mechanisms of plant morphogenesis.

PSAC-PDB: Analysis and classification of protein structures

This paper presents a novel framework, called PSAC-PDB, for analyzing and classifying protein structures from the Protein Data Bank (PDB). PSAC-PDB first finds, analyze and identifies protein structures in PDB that are similar to a protein structure of interest using a protein structure comparison tool. Second, the amino acids (AA) sequences of identified protein structures (obtained from PDB), their aligned amino acids (AAA) and aligned secondary structure elements (ASSE) (obtained by structural alignment), and frequent AA (FAA) patterns (discovered by sequential pattern mining), are used for the reliable detection/classification of protein structures. Eleven classifiers are used and their performance is compared using six evaluation metrics. Results show that three classifiers perform well on overall, and that FAA patterns can be used to efficiently classify protein structures in place of providing the whole AA sequences, AAA or ASSE. Furthermore, better classification results are obtained using AAA of protein structures rather than AA sequences. PSAC-PDB also performed better than state-of-the-art approaches for SARS-CoV-2 genome sequences classification.

DomBpred: Protein Domain Boundary Prediction Based on Domain-Residue Clustering Using Inter-Residue Distance

Domain boundary prediction is one of the most important problems in the study of protein structure and function, especially for large proteins. At present, most domain boundary prediction methods have low accuracy and limitations in dealing with multi-domain proteins. In this study, we develop a sequence-based protein domain boundary prediction, named DomBpred. In DomBpred, the input sequence is first classified as either a single-domain protein or a multi-domain protein through a designed effective sequence metric based on a constructed single-domain sequence library. For the multi-domain protein, a domain-residue clustering algorithm inspired by Ising model is proposed to cluster the spatially close residues according inter-residue distance. The unclassified residues and the residues at the edge of the cluster are then tuned by the secondary structure to form potential cut points. Finally, a domain boundary scoring function is proposed to recursively evaluate the potential cut points to generate the domain boundary. DomBpred is tested on a large-scale test set of FUpred comprising 2549 proteins. Experimental results show that DomBpred better performs than the state-of-the-art methods in classifying whether protein sequences are composed by single or multiple domains, and the Matthew's correlation coefficient is 0.882. Moreover, on 849 multi-domain proteins, the domain boundary distance and normalised domain overlap scores of DomBpred are 0.523 and 0.824, respectively, which are 5.0% and 4.2% higher than those of the best comparison method, respectively. Comparison with other methods on the given test set shows that DomBpred outperforms most state-of-the-art sequence-based methods and even achieves better results than the top-level template-based method. The executable program is freely available at https://github.com/iobio-zjut/DomBpred and the online server at http://zhanglab-bioinf.com/DomBpred/.

A novel strategy for designing the magic shotguns for distantly related target pairs

Due to its promising capacity in improving drug efficacy, polypharmacology has emerged to be a new theme in the drug discovery of complex disease. In the process of novel multi-target drugs (MTDs) discovery, in silico strategies come to be quite essential for the advantage of high throughput and low cost. However, current researchers mostly aim at typical closely related target pairs. Because of the intricate pathogenesis networks of complex diseases, many distantly related targets are found to play crucial role in synergistic treatment. Therefore, an innovational method to develop drugs which could simultaneously target distantly related target pairs is of utmost importance. At the same time, reducing the false discovery rate in the design of MTDs remains to be the daunting technological difficulty. In this research, effective small molecule clustering in the positive dataset, together with a putative negative dataset generation strategy, was adopted in the process of model constructions. Through comprehensive assessment on 10 target pairs with hierarchical similarity-levels, the proposed strategy turned out to reduce the false discovery rate successfully. Constructed model types with much smaller numbers of inhibitor molecules gained considerable yields and showed better false-hit controllability than before. To further evaluate the generalization ability, an in-depth assessment of high-throughput virtual screening on ChEMBL database was conducted. As a result, this novel strategy could hierarchically improve the enrichment factors for each target pair (especially for those distantly related/unrelated target pairs), corresponding to target pair similarity-levels.

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.

Deformable Protein Shape Classification Based on Deep Learning, and the Fractional Fokker-Planck and Kähler-Dirac Equations

The classification of deformable protein shapes, based solely on their macromolecular surfaces, is a challenging problem in protein-protein interaction prediction and protein design. Shape classification is made difficult by the fact that proteins are dynamic, flexible entities with high geometrical complexity. In this paper, we introduce a novel description for such deformable shapes. This description is based on the bifractional Fokker-Planck and Dirac-Kähler equations. These equations analyse and probe protein shapes in terms of a scalar, vectorial and non-commuting quaternionic field, allowing for a more comprehensive description of the protein shapes. An underlying non-Markovian Lévy random walk establishes geometrical relationships between distant regions while recalling previous analyses. Classification is performed with a multiobjective deep hierarchical pyramidal neural network, thus performing a multilevel analysis of the description. Our approach is applied to the SHREC'19 dataset for deformable protein shapes classification and to the SHREC'16 dataset for deformable partial shapes classification, demonstrating the effectiveness and generality of our approach.

The COMER web server for protein analysis by homology

SUMMARY

Sequence homology is a basic concept in protein evolution, structure and function studies. However, there are not many different tools and services for homology searches being sensitive, accurate and fast at the same time. We present a new web server for protein analysis based on COMER2, a sequence alignment and homology search method that exhibits these characteristics. COMER2 has been upgraded since its last publication to improve its alignment quality and ease of use. We demonstrate how the user can benefit from using it by providing examples of extensive annotation of proteins of unknown function. Among the distinctive features of the web server is the user's ability to submit multiple queries with one click of a button. This and other features allow for transparently running homology searches-in a command-line, programmatic or graphical environment-across multiple databases with multiple queries. They also promote extensive simultaneous protein analysis at the sequence, structure and function levels.

AVAILABILITY AND IMPLEMENTATION

The COMER web server is available at https://bioinformatics.lt/comer.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Intrinsically disordered regions that drive phase separation form a robustly distinct protein class

Protein phase separation is thought to be a primary driving force for the formation of membrane-less organelles, which control a wide range of biological functions from stress response to ribosome biogenesis. Among phase-separating (PS) proteins, many have intrinsically disordered regions (IDRs) that are needed for phase separation to occur. Accurate identification of IDRs that drive phase separation is important for testing the underlying mechanisms of phase separation, identifying biological processes that rely on phase separation, and designing sequences that modulate phase separation. To identify IDRs that drive phase separation, we first curated datasets of folded, ID, and PS ID sequences. We then used these sequence sets to examine how broadly existing amino acid property scales can be used to distinguish between the three classes of protein regions. We found that there are robust property differences between the classes and, consequently, that numerous combinations of amino acid property scales can be used to make robust predictions of protein phase separation. This result indicates that multiple, redundant mechanisms contribute to the formation of phase-separated droplets from IDRs. The top-performing scales were used to further optimize our previously developed predictor of PS IDRs, ParSe. We then modified ParSe to account for interactions between amino acids and obtained reasonable predictive power for mutations that have been designed to test the role of amino acid interactions in driving protein phase separation. Collectively, our findings provide further insight into the classification of IDRs and the elements involved in protein phase separation.

RCSB Protein Data bank: Tools for visualizing and understanding biological macromolecules in 3D

Now in its 52nd year of continuous operations, the Protein Data Bank (PDB) is the premiere open-access global archive housing three-dimensional (3D) biomolecular structure data. It is jointly managed by the Worldwide Protein Data Bank (wwPDB) partnership. The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB) is funded by the National Science Foundation, National Institutes of Health, and US Department of Energy and serves as the US data center for the wwPDB. RCSB PDB is also responsible for the security of PDB data in its role as wwPDB-designated Archive Keeper. Every year, RCSB PDB serves tens of thousands of depositors of 3D macromolecular structure data (coming from macromolecular crystallography, nuclear magnetic resonance spectroscopy, electron microscopy, and micro-electron diffraction). The RCSB PDB research-focused web portal (RCSB.org) makes PDB data available at no charge and without usage restrictions to many millions of PDB data consumers around the world. The RCSB PDB training, outreach, and education web portal (PDB101.RCSB.org) serves nearly 700 K educators, students, and members of the public worldwide. This invited Tools Issue contribution describes how RCSB PDB (i) is organized; (ii) works with wwPDB partners to process new depositions; (iii) serves as the wwPDB-designated Archive Keeper; (iv) enables exploration and 3D visualization of PDB data via RCSB.org; and (v) supports training, outreach, and education via PDB101.RCSB.org. New tools and features at RCSB.org are presented using examples drawn from high-resolution structural studies of proteins relevant to treatment of human cancers by targeting immune checkpoints.

Statistical proofs of the interdependence between nearest neighbor effects on polypeptide backbone conformations

Backbone dihedral angles ϕ and ψ are the main structural descriptors of proteins and peptides. The distribution of these angles has been investigated over decades as they are essential for the validation and refinement of experimental measurements, as well as for structure prediction and design methods. The dependence of these distributions, not only on the nature of each amino acid but also on that of the closest neighbors, has been the subject of numerous studies. Although neighbor-dependent distributions are nowadays generally accepted as a good model, there is still some controversy about the combined effects of left and right neighbors. We have investigated this question using rigorous methods based on recently-developed statistical techniques. Our results unambiguously demonstrate that the influence of left and right neighbors cannot be considered independently. Consequently, three-residue fragments should be considered as the minimal building blocks to investigate polypeptide sequence-structure relationships.

The Venturia inaequalis effector repertoire is dominated by expanded families with predicted structural similarity, but unrelated sequence, to avirulence proteins from other plant-pathogenic fungi

Background

Scab, caused by the biotrophic fungus Venturia inaequalis, is the most economically important disease of apples worldwide. During infection, V. inaequalis occupies the subcuticular environment, where it secretes virulence factors, termed effectors, to promote host colonization. Consistent with other plant-pathogenic fungi, many of these effectors are expected to be non-enzymatic proteins, some of which can be recognized by corresponding host resistance proteins to activate plant defences, thus acting as avirulence determinants. To develop durable control strategies against scab, a better understanding of the roles that these effector proteins play in promoting subcuticular growth by V. inaequalis, as well as in activating, suppressing, or circumventing resistance protein-mediated defences in apple, is required.

Results

We generated the first comprehensive RNA-seq transcriptome of V. inaequalis during colonization of apple. Analysis of this transcriptome revealed five temporal waves of gene expression that peaked during early, mid, or mid-late infection. While the number of genes encoding secreted, non-enzymatic proteinaceous effector candidates (ECs) varied in each wave, most belonged to waves that peaked in expression during mid-late infection. Spectral clustering based on sequence similarity determined that the majority of ECs belonged to expanded protein families. To gain insights into function, the tertiary structures of ECs were predicted using AlphaFold2. Strikingly, despite an absence of sequence similarity, many ECs were predicted to have structural similarity to avirulence proteins from other plant-pathogenic fungi, including members of the MAX, LARS, ToxA and FOLD effector families. In addition, several other ECs, including an EC family with sequence similarity to the AvrLm6 avirulence effector from Leptosphaeria maculans, were predicted to adopt a KP6-like fold. Thus, proteins with a KP6-like fold represent another structural family of effectors shared among plant-pathogenic fungi.

Conclusions

Our study reveals the transcriptomic profile underpinning subcuticular growth by V. inaequalis and provides an enriched list of ECs that can be investigated for roles in virulence and avirulence. Furthermore, our study supports the idea that numerous sequence-unrelated effectors across plant-pathogenic fungi share common structural folds. In doing so, our study gives weight to the hypothesis that many fungal effectors evolved from ancestral genes through duplication, followed by sequence diversification, to produce sequence-unrelated but structurally similar proteins.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-022-01442-9.

A family of unusual immunoglobulin superfamily genes in an invertebrate histocompatibility complex

Most colonial marine invertebrates are capable of allorecognition, the ability to distinguish between themselves and conspecifics. One long-standing question is whether invertebrate allorecognition genes are homologous to vertebrate histocompatibility genes. In the cnidarian Hydractinia symbiolongicarpus, allorecognition is controlled by at least two genes, Allorecognition 1 (Alr1) and Allorecognition 2 (Alr2), which encode highly polymorphic cell-surface proteins that serve as markers of self. Here, we show that Alr1 and Alr2 are part of a family of 41 Alr genes, all of which reside in a single genomic interval called the Allorecognition Complex (ARC). Using sensitive homology searches and highly accurate structural predictions, we demonstrate that the Alr proteins are members of the immunoglobulin superfamily (IgSF) with V-set and I-set Ig domains unlike any previously identified in animals. Specifically, their primary amino acid sequences lack many of the motifs considered diagnostic for V-set and I-set domains, yet they adopt secondary and tertiary structures nearly identical to canonical Ig domains. Thus, the V-set domain, which played a central role in the evolution of vertebrate adaptive immunity, was present in the last common ancestor of cnidarians and bilaterians. Unexpectedly, several Alr proteins also have immunoreceptor tyrosine-based activation motifs and immunoreceptor tyrosine-based inhibitory motifs in their cytoplasmic tails, suggesting they could participate in pathways homologous to those that regulate immunity in humans and flies. This work expands our definition of the IgSF with the addition of a family of unusual members, several of which play a role in invertebrate histocompatibility.

Protein Data Bank: A Comprehensive Review of 3D Structure Holdings and Worldwide Utilization by Researchers, Educators, and Students

The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB), funded by the United States National Science Foundation, National Institutes of Health, and Department of Energy, supports structural biologists and Protein Data Bank (PDB) data users around the world. The RCSB PDB, a founding member of the Worldwide Protein Data Bank (wwPDB) partnership, serves as the US data center for the global PDB archive housing experimentally-determined three-dimensional (3D) structure data for biological macromolecules. As the wwPDB-designated Archive Keeper, RCSB PDB is also responsible for the security of PDB data and weekly update of the archive. RCSB PDB serves tens of thousands of data depositors (using macromolecular crystallography, nuclear magnetic resonance spectroscopy, electron microscopy, and micro-electron diffraction) annually working on all permanently inhabited continents. RCSB PDB makes PDB data available from its research-focused web portal at no charge and without usage restrictions to many millions of PDB data consumers around the globe. It also provides educators, students, and the general public with an introduction to the PDB and related training materials through its outreach and education-focused web portal. This review article describes growth of the PDB, examines evolution of experimental methods for structure determination viewed through the lens of the PDB archive, and provides a detailed accounting of PDB archival holdings and their utilization by researchers, educators, and students worldwide.

ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

Assignment of structural domains in proteins using diffusion kernels on graphs

Though proposing algorithmic approaches for protein domain decomposition has been of high interest, the inherent ambiguity to the problem makes it still an active area of research. Besides, accurate automated methods are in high demand as the number of solved structures for complex proteins is on the rise. While majority of the previous efforts for decomposition of 3D structures are centered on the developing clustering algorithms, employing enhanced measures of proximity between the amino acids has remained rather uncharted. If there exists a kernel function that in its reproducing kernel Hilbert space, structural domains of proteins become well separated, then protein structures can be parsed into domains without the need to use a complex clustering algorithm. Inspired by this idea, we developed a protein domain decomposition method based on diffusion kernels on protein graphs. We examined all combinations of four graph node kernels and two clustering algorithms to investigate their capability to decompose protein structures. The proposed method is tested on five of the most commonly used benchmark datasets for protein domain assignment plus a comprehensive non-redundant dataset. The results show a competitive performance of the method utilizing one of the diffusion kernels compared to four of the best automatic methods. Our method is also able to offer alternative partitionings for the same structure which is in line with the subjective definition of protein domain. With a competitive accuracy and balanced performance for the simple and complex structures despite relying on a relatively naive criterion to choose optimal decomposition, the proposed method revealed that diffusion kernels on graphs in particular, and kernel functions in general are promising measures to facilitate parsing proteins into domains and performing different structural analysis on proteins. The size and interconnectedness of the protein graphs make them promising targets for diffusion kernels as measures of affinity between amino acids. The versatility of our method allows the implementation of future kernels with higher performance. The source code of the proposed method is accessible at https://github.com/taherimo/kludo . Also, the proposed method is available as a web application from https://cbph.ir/tools/kludo .

ASFold-DNN: Protein Fold Recognition Based on Evolutionary Features With Variable Parameters Using Full Connected Neural Network

Protein fold recognition contribute to comprehend the function of proteins, which is of great help to the gene therapy of diseases and the development of new drugs. Researchers have been working in this direction and have made considerable achievements, but challenges still exist on low sequence similarity datasets. In this study, we propose the ASFold-DNN framework for protein fold recognition research. Above all, four groups of evolutionary features are extracted from the primary structures of proteins, and a preliminary selection of variable parameter is made for two groups of features including ACC _HMM and SXG _HMM, respectively. Then several feature selection algorithms are selected for comparison and the best feature selection scheme is obtained by changing their internal threshold values. Finally, multiple hyper-parameters of Full Connected Neural Network are fully optimized to construct the best model. DD, EDD and TG datasets with low sequence similarities are chosen to evaluate the performance of the models constructed by the framework, and the final prediction accuracy are 85.28, 95.00 and 88.84 percent, respectively. Furthermore, the ASTRAL186 and LE datasets are introduced to further verify the generalization ability of our proposed framework. Comprehensive experimental results prove that the ASFold-DNN framework is more prominent than the state-of-the-art studies on protein fold recognition. The source code and data of ASFold-DNN can be downloaded from https://github.com/Bioinformatics-Laboratory/project/tree/master/ASFold.

Multi-layer sequential network analysis improves protein 3D structural classification

Protein structural classification (PSC) is a supervised problem of assigning proteins into pre-defined structural (e.g., CATH or SCOPe) classes based on the proteins' sequence or 3D structural features. We recently proposed PSC approaches that model protein 3D structures as protein structure networks (PSNs) and analyze PSN-based protein features, which performed better than or comparable to state-of-the-art sequence or other 3D structure-based PSC approaches. However, existing PSN-based PSC approaches model the whole 3D structure of a protein as a static (i.e., single-layer) PSN. Because folding of a protein is a dynamic process, where some parts (i.e., sub-structures) of a protein fold before others, modeling the 3D structure of a protein as a PSN that captures the sub-structures might further help improve the existing PSC performance. Here, we propose to model 3D structures of proteins as multi-layer sequential PSNs that approximate 3D sub-structures of proteins, with the hypothesis that this will improve upon the current state-of-the-art PSC approaches that are based on single-layer PSNs (and thus upon the existing state-of-the-art sequence and other 3D structural approaches). Indeed, we confirm this on 72 datasets spanning ~44 000 CATH and SCOPe protein domains.

Low Complexity Induces Structure in Protein Regions Predicted as Intrinsically Disordered

There is increasing evidence that many intrinsically disordered regions (IDRs) in proteins play key functional roles through interactions with other proteins or nucleic acids. These interactions often exhibit a context-dependent structural behavior. We hypothesize that low complexity regions (LCRs), often found within IDRs, could have a role in inducing local structure in IDRs. To test this, we predicted IDRs in the human proteome and analyzed their structures or those of homologous sequences in the Protein Data Bank (PDB). We then identified two types of simple LCRs within IDRs: regions with only one (polyX or homorepeats) or with only two types of amino acids (polyXY). We were able to assign structural information from the PDB more often to these LCRs than to the surrounding IDRs (polyX 61.8% > polyXY 50.5% > IDRs 39.7%). The most frequently observed polyX and polyXY within IDRs contained E (Glu) or G (Gly). Structural analyses of these sequences and of homologs indicate that polyEK regions induce helical conformations, while the other most frequent LCRs induce coil structures. Our work proposes bioinformatics methods to help in the study of the structural behavior of IDRs and provides a solid basis suggesting a structuring role of LCRs within them.

Vexitoxins: conotoxin-like venom peptides from predatory gastropods of the genus Vexillum

Venoms of predatory marine cone snails are intensely studied because of the biomedical applications of the neuropeptides that they contain, termed conotoxins. Meanwhile some gastropod lineages have independently acquired secretory glands strikingly similar to the venom gland of cone snails, suggesting that they possess similar venoms. Here we focus on the most diversified of these clades, the genus Vexillum. Based on the analysis of a multi-species proteo-transcriptomic dataset, we show that Vexillum species indeed produce complex venoms dominated by highly diversified short cysteine-rich peptides, vexitoxins. Vexitoxins possess the same precursor organization, display overlapping cysteine frameworks and share several common post-translational modifications with conotoxins. Some vexitoxins show sequence similarity to conotoxins and adopt similar domain conformations, including a pharmacologically relevant inhibitory cysteine knot motif. The Vexillum envenomation gland (gL) is a notably more recent evolutionary novelty than the conoidean venom gland. Thus, we hypothesize lower divergence between vexitoxin genes, and their ancestral 'somatic' counterparts compared to that in conotoxins, and we find support for this hypothesis in the evolution of the vexitoxin cluster V027. We use this example to discuss how future studies on vexitoxins can inform the origin of conotoxins, and how they may help to address outstanding questions in venom evolution.

A closed Candidatus Odinarchaeum chromosome exposes Asgard archaeal viruses

Asgard archaea have recently been identified as the closest archaeal relatives of eukaryotes. Their ecology, and particularly their virome, remain enigmatic. We reassembled and closed the chromosome of Candidatus Odinarchaeum yellowstonii LCB_4, through long-range PCR, revealing CRISPR spacers targeting viral contigs. We found related viruses in the genomes of diverse prokaryotes from geothermal environments, including other Asgard archaea. These viruses open research avenues into the ecology and evolution of Asgard archaea.

Edge weights in a protein elastic network reorganize collective motions and render long-range sensitivity responses

The effects of inter-residue interactions on protein collective motions are analyzed by comparing two elastic network models (ENM)—structural contact ENM (SC-ENM) and molecular dynamics (MD)-ENM—with the edge weights computed from an all-atom MD trajectory by structure-mechanics statistical learning. A theoretical framework is devised to decompose the eigenvalues of ENM Hessian into contributions from individual springs and to compute the sensitivities of positional fluctuations and covariances to spring constant variation. Our linear perturbation approach quantifies the response mechanisms as softness modulation and orientation shift. All contacts of C α positions in SC-ENM have an identical spring constant by fitting the profile of root-of-mean-squared-fluctuation calculated from an all-atom MD simulation, and the same trajectory data are also used to compute the specific spring constant of each contact as an MD-ENM edge weight. We illustrate that the soft-mode reorganization can be understood in terms of gaining weights along the structural contacts of low elastic strengths and loosing magnitude along those of high rigidities. With the diverse mechanical strengths encoded in protein dynamics, MD-ENM is found to have more pronounced long-range couplings and sensitivity responses with orientation shift identified as a key player in driving the specific residues to have high sensitivities. Furthermore, the responses of perturbing the springs of different residues are found to have asymmetry in the action–reaction relationship. In understanding the mutation effects on protein functional properties, such as long-range communications, our results point in the directions of collective motions as a major effector.

The Structural Rule Distinguishing a Superfold: A Case Study of Ferredoxin Fold and the Reverse Ferredoxin Fold

Superfolds are folds commonly observed among evolutionarily unrelated multiple superfamilies of proteins. Since discovering superfolds almost two decades ago, structural rules distinguishing superfolds from the other ordinary folds have been explored but remained elusive. Here, we analyzed a typical superfold, the ferredoxin fold, and the fold which reverses the N to C terminus direction from the ferredoxin fold as a case study to find the rule to distinguish superfolds from the other folds. Though all the known structural characteristics for superfolds apply to both the ferredoxin fold and the reverse ferredoxin fold, the reverse fold has been found only in a single superfamily. The database analyses in the present study revealed the structural preferences of αβ- and βα-units; the preferences separate two α-helices in the ferredoxin fold, preventing their collision and stabilizing the fold. In contrast, in the reverse ferredoxin fold, the preferences bring two helices near each other, inducing structural conflict. The Rosetta folding simulations suggested that the ferredoxin fold is physically much more realizable than the reverse ferredoxin fold. Therefore, we propose that minimal structural conflict or minimal frustration among secondary structures is the rule to distinguish a superfold from ordinary folds. Intriguingly, the database analyses revealed that a most stringent structural rule in proteins, the right-handedness of the βαβ-unit, is broken in a set of structures to prevent the frustration, suggesting the proposed rule of minimum frustration among secondary structural units is comparably strong as the right-handedness rule of the βαβ-unit.

Dali server: structural unification of protein families

Protein structure is key to understanding biological function. Structure comparison deciphers deep phylogenies, providing insight into functional conservation and functional shifts during evolution. Until recently, structural coverage of the protein universe was limited by the cost and labour involved in experimental structure determination. Recent breakthroughs in deep learning revolutionized structural bioinformatics by providing accurate structural models of numerous protein families for which no structural information existed. The Dali server for 3D protein structure comparison is widely used by crystallographers to relate new structures to pre-existing ones. Here, we report two most recent upgrades to the web server: (i) the foldomes of key organisms in the AlphaFold Database (version 1) are searchable by Dali, (ii) structural alignments are annotated with protein families. Using these new features, we discovered a novel functionally diverse subgroup within the WRKY/GCM1 clan. This was accomplished by linking the structurally characterized SWI/SNF and NAM families as well as the structural models of the CG-1 family and uncharacterized proteins to the structure of Gti1/Pac2, a previously known member of the WRKY/GCM1 clan. The Dali server is available at http://ekhidna2.biocenter.helsinki.fi/dali. This website is free and open to all users and there is no login requirement.

The Human Ntn-Hydrolase Superfamily: Structure, Functions and Perspectives

N-terminal nucleophile (Ntn)-hydrolases catalyze the cleavage of amide bonds in a variety of macromolecules, including the peptide bond in proteins, the amide bond in N-linked protein glycosylation, and the amide bond linking a fatty acid to sphingosine in complex sphingolipids. Ntn-hydrolases are all sharing two common hallmarks: Firstly, the enzymes are synthesized as inactive precursors that undergo auto-proteolytic self-activation, which, as a consequence, reveals the active site nucleophile at the newly formed N-terminus. Secondly, all Ntn-hydrolases share a structural consistent αββα-fold, notwithstanding the total lack of amino acid sequence homology. In humans, five subclasses of the Ntn-superfamily have been identified so far, comprising relevant members such as the catalytic active subunits of the proteasome or a number of lysosomal hydrolases, which are often associated with lysosomal storage diseases. This review gives an updated overview on the structural, functional, and (patho-)physiological characteristics of human Ntn-hydrolases, in particular.

LOMETS3: integrating deep learning and profile alignment for advanced protein template recognition and function annotation

Deep learning techniques have significantly advanced the field of protein structure prediction. LOMETS3 (https://zhanglab.ccmb.med.umich.edu/LOMETS/) is a new generation meta-server approach to template-based protein structure prediction and function annotation, which integrates newly developed deep learning threading methods. For the first time, we have extended LOMETS3 to handle multi-domain proteins and to construct full-length models with gradient-based optimizations. Starting from a FASTA-formatted sequence, LOMETS3 performs four steps of domain boundary prediction, domain-level template identification, full-length template/model assembly and structure-based function prediction. The output of LOMETS3 contains (i) top-ranked templates from LOMETS3 and its component threading programs, (ii) up to 5 full-length structure models constructed by L-BFGS (limited-memory Broyden–Fletcher–Goldfarb–Shanno algorithm) optimization, (iii) the 10 closest Protein Data Bank (PDB) structures to the target, (iv) structure-based functional predictions, (v) domain partition and assembly results, and (vi) the domain-level threading results, including items (i)–(iii) for each identified domain. LOMETS3 was tested in large-scale benchmarks and the blind CASP14 (14th Critical Assessment of Structure Prediction) experiment, where the overall template recognition and function prediction accuracy is significantly beyond its predecessors and other state-of-the-art threading approaches, especially for hard targets without homologous templates in the PDB. Based on the improved developments, LOMETS3 should help significantly advance the capability of broader biomedical community for template-based protein structure and function modelling.

Real-time structure search and structure classification for AlphaFold protein models

Last year saw a breakthrough in protein structure prediction, where the AlphaFold2 method showed a substantial improvement in the modeling accuracy. Following the software release of AlphaFold2, predicted structures by AlphaFold2 for proteins in 21 species were made publicly available via the AlphaFold Database. Here, to facilitate structural analysis and application of AlphaFold2 models, we provide the infrastructure, 3D-AF-Surfer, which allows real-time structure-based search for the AlphaFold2 models. In 3D-AF-Surfer, structures are represented with 3D Zernike descriptors (3DZD), which is a rotationally invariant, mathematical representation of 3D shapes. We developed a neural network that takes 3DZDs of proteins as input and retrieves proteins of the same fold more accurately than direct comparison of 3DZDs. Using 3D-AF-Surfer, we report structure classifications of AlphaFold2 models and discuss the correlation between confidence levels of AlphaFold2 models and intrinsic disordered regions.

Understanding structural variability in proteins using protein structural networks

Proteins perform their function by accessing a suitable conformer from the ensemble of available conformations. The conformational diversity of a chosen protein structure can be obtained by experimental methods under different conditions. A key issue is the accurate comparison of different conformations. A gold standard used for such a comparison is the root mean square deviation (RMSD) between the two structures. While extensive refinements of RMSD evaluation at the backbone level are available, a comprehensive framework including the side chain interaction is not well understood. Here we employ protein structure network (PSN) formalism, with the non-covalent interactions of side chain, explicitly treated. The PSNs thus constructed are compared through graph spectral method, which provides a comparison at the local and at the global structural level. In this work, PSNs of multiple crystal conformers of single-chain, single-domain proteins, are subject to pair-wise analysis to examine the dissimilarity in their network topologies and in order to determine the conformational diversity of their native structures. This information is utilized to classify the structural domains of proteins into different categories. It is observed that proteins typically tend to retain structure and interactions at the backbone level. However, some of them also depict variability in either their overall structure or only in their inter-residue connectivity at the sidechain level, or both. Variability of sub-networks based on solvent accessibility and secondary structure is studied. The types of specific interactions are found to contribute differently to structure variability. An ensemble analysis by computing the mathematical variance of edge-weights across multiple conformers provided information on the contribution to overall variability from each edge of the PSN. Interactions that are highly variable are identified and their impact on structure variability has been discussed with the help of a case study. The classification based on the present side-chain network-based studies provides a framework to correlate the structure-function relationships in protein structures.

•

Monomeric, single domain protein structures can exhibit non-rigid behaviour and be highly variable.

•

The comparison of protein structural networks can better discriminate conformations with similar backbones.

•

Specific interactions between solvent accessible and inaccessible residues are poorly preserved.

•

Network edge-variation offers insights on which interacting residues are likely to influence their dynamics and function.

•

These side-chain network-based studies provide a framework to correlate protein structure-function relationships.

Molecular Modeling is an Enabling Approach to Complement and Enhance Channelopathy Research

Hundreds of human membrane proteins form channels that transport necessary ions and compounds, including drugs and metabolites, yet details of their normal function or how function is altered by genetic variants to cause diseases are often unknown. Without this knowledge, researchers are less equipped to develop approaches to diagnose and treat channelopathies. High-resolution computational approaches such as molecular modeling enable researchers to investigate channelopathy protein function, facilitate detailed hypothesis generation, and produce data that is difficult to gather experimentally. Molecular modeling can be tailored to each physiologic context that a protein may act within, some of which may currently be difficult or impossible to assay experimentally. Because many genomic variants are observed in channelopathy proteins from high-throughput sequencing studies, methods with mechanistic value are needed to interpret their effects. The eminent field of structural bioinformatics integrates techniques from multiple disciplines including molecular modeling, computational chemistry, biophysics, and biochemistry, to develop mechanistic hypotheses and enhance the information available for understanding function. Molecular modeling and simulation access 3D and time-dependent information, not currently predictable from sequence. Thus, molecular modeling is valuable for increasing the resolution with which the natural function of protein channels can be investigated, and for interpreting how genomic variants alter them to produce physiologic changes that manifest as channelopathies. © 2022 American Physiological Society. Compr Physiol 12:3141-3166, 2022.

Surface-based protein domains retrieval methods from a SHREC2021 challenge

Proteins are essential to nearly all cellular mechanism and the effectors of the cells activities. As such, they often interact through their surface with other proteins or other cellular ligands such as ions or organic molecules. The evolution generates plenty of different proteins, with unique abilities, but also proteins with related functions hence similar 3D surface properties (shape, physico-chemical properties, …). The protein surfaces are therefore of primary importance for their activity. In the present work, we assess the ability of different methods to detect such similarities based on the geometry of the protein surfaces (described as 3D meshes), using either their shape only, or their shape and the electrostatic potential (a biologically relevant property of proteins surface). Five different groups participated in this contest using the shape-only dataset, and one group extended its pre-existing method to handle the electrostatic potential. Our comparative study reveals both the ability of the methods to detect related proteins and their difficulties to distinguish between highly related proteins. Our study allows also to analyze the putative influence of electrostatic information in addition to the one of protein shapes alone. Finally, the discussion permits to expose the results with respect to ones obtained in the previous contests for the extended method. The source codes of each presented method have been made available online.

Membrane contact probability: An essential and predictive character for the structural and functional studies of membrane proteins

One of the unique traits of membrane proteins is that a significant fraction of their hydrophobic amino acids is exposed to the hydrophobic core of lipid bilayers rather than being embedded in the protein interior, which is often not explicitly considered in the protein structure and function predictions. Here, we propose a characteristic and predictive quantity, the membrane contact probability (MCP), to describe the likelihood of the amino acids of a given sequence being in direct contact with the acyl chains of lipid molecules. We show that MCP is complementary to solvent accessibility in characterizing the outer surface of membrane proteins, and it can be predicted for any given sequence with a machine learning-based method by utilizing a training dataset extracted from MemProtMD, a database generated from molecular dynamics simulations for the membrane proteins with a known structure. As the first of many potential applications, we demonstrate that MCP can be used to systematically improve the prediction precision of the protein contact maps and structures.

The distribution of residues on protein surfaces is largely determined by the surrounding environment. For soluble proteins, most of the residues on the outer surface are hydrophilic, and people use the quantity “solvent accessibility” to describe and predict these surface residues. In contrast, for membrane proteins that are embedded in a lipid bilayer, many of their surface residues are hydrophobic and membrane-contacting, but there is yet a widely-accepted quantity for the description or prediction of this characteristic property. Here, we propose a new quantity termed “membrane contact probability (MCP)”, which can be used to describe and predict the membrane-contacting surface residues of proteins. We also propose a machine learning-based method to predict MCP from protein sequences, utilizing the dataset generated by physics-based computer simulations. We demonstrate that a quantity such as MCP is helpful for protein structure prediction, and we believe that it will find broad applications in the structure and function studies of membrane proteins.

Mining folded proteomes in the era of accurate structure prediction

Protein structure fundamentally underpins the function and processes of numerous biological systems. Fold recognition algorithms offer a sensitive and robust tool to detect structural, and thereby functional, similarities between distantly related homologs. In the era of accurate structure prediction owing to advances in machine learning techniques and a wealth of experimentally determined structures, previously curated sequence databases have become a rich source of biological information. Here, we use bioinformatic fold recognition algorithms to scan the entire AlphaFold structure database to identify novel protein family members, infer function and group predicted protein structures. As an example of the utility of this approach, we identify novel, previously unknown members of various pore-forming protein families, including MACPFs, GSDMs and aerolysin-like proteins.

Virtually every cellular process in all organisms on Earth is driven by molecular nano-machines known as proteins. The diverse functions of proteins are the result of the unique three-dimensional shape adopted by a given protein molecule. It is therefore important to determine the shape of a given protein, which unlike DNA and our genes, cannot be known from its sequence alone. Since two proteins with similar shapes typically have a similar function, knowing a protein shape provides crucial clues about its function. By virtue of decades of experimental work and advances in artificial intelligence, this complex shape can now be computationally predicted for any protein whose composition is known. Scientists have used these and other methods to produce enormous libraries of protein shapes consisting of nearly a million unique entries. However, these libraries are too large and too complex for researchers to ‘read’. We use shape-comparison algorithms to carefully check these shape-libraries to gain insight into the potential function and biological role of previously unknown proteins. Furthermore, we identified new members of protein families using this technique. We show that shape-matching algorithms and computationally generated shape-libraries can be used effectively together to yield new insights and expedite scientific endeavours.

Quantitative analysis of visual codewords of a protein distance matrix

3D protein structures can be analyzed using a distance matrix calculated as the pairwise distance between all Cα atoms in the protein model. Although researchers have efficiently used distance matrices to classify proteins and find homologous proteins, much less work has been done on quantitative analysis of distance matrix features. Therefore, the distance matrix was analyzed as gray scale image using KAZE feature extractor algorithm with Bag of Visual Words model. In this study, each protein was represented as a histogram of visual codewords. The analysis showed that a very small number of codewords (~1%) have a high relative frequency (> 0.25) and that the majority of codewords have a relative frequency around 0.05. We have also shown that there is a relationship between the frequency of codewords and the position of the features in a distance matrix. The codewords that are more frequent are located closer to the main diagonal. Less frequent codewords, on the other hand, are located in the corners of the distance matrix, far from the main diagonal. Moreover, the analysis showed a correlation between the number of unique codewords and the 3D repeats in the protein structure. The solenoid and tandem repeats proteins have a significantly lower number of unique codewords than the globular proteins. Finally, the codeword histograms and Support Vector Machine (SVM) classifier were used to classify solenoid and globular proteins. The result showed that the SVM classifier fed with codeword histograms correctly classified 352 out of 354 proteins.

PHR-search: a search framework for protein remote homology detection based on the predicted protein hierarchical relationships

Protein remote homology detection is one of the most fundamental research tool for protein structure and function prediction. Most search methods for protein remote homology detection are evaluated based on the Structural Classification of Proteins-extended (SCOPe) benchmark, but the diverse hierarchical structure relationships between the query protein and candidate proteins are ignored by these methods. In order to further improve the predictive performance for protein remote homology detection, a search framework based on the predicted protein hierarchical relationships (PHR-search) is proposed. In the PHR-search framework, the superfamily level prediction information is obtained by extracting the local and global features of the Hidden Markov Model (HMM) profile through a convolution neural network and it is converted to the fold level and class level prediction information according to the hierarchical relationships of SCOPe. Based on these predicted protein hierarchical relationships, filtering strategy and re-ranking strategy are used to construct the two-level search of PHR-search. Experimental results show that the PHR-search framework achieves the state-of-the-art performance by employing five basic search methods, including HHblits, JackHMMER, PSI-BLAST, DELTA-BLAST and PSI-BLASTexB. Furthermore, the web server of PHR-search is established, which can be accessed at http://bliulab.net/PHR-search.

Interpreting Potts and Transformer Protein Models Through the Lens of Simplified Attention

The established approach to unsupervised protein contact prediction estimates coevolving positions using undirected graphical models. This approach trains a Potts model on a Multiple Sequence Alignment. Increasingly large Transformers are being pretrained on unlabeled, unaligned protein sequence databases and showing competitive performance on protein contact prediction. We argue that attention is a principled model of protein interactions, grounded in real properties of protein family data. We introduce an energy-based attention layer, factored attention, which, in a certain limit, recovers a Potts model, and use it to contrast Potts and Transformers. We show that the Transformer leverages hierarchical signal in protein family databases not captured by single-layer models. This raises the exciting possibility for the development of powerful structured models of protein family databases.

ProFitFun: a protein tertiary structure fitness function for quantifying the accuracies of model structures

MOTIVATION

An accurate estimation of the quality of protein model structures typifies as a cornerstone in protein structure prediction regimes. Despite the recent groundbreaking success in the field of protein structure prediction, there are certain prospects for the improvement in model quality estimation at multiple stages of protein structure prediction and thus, to further push the prediction accuracy. Here, a novel approach, named ProFitFun, for assessing the quality of protein models is proposed by harnessing the sequence and structural features of experimental protein structures in terms of the preferences of backbone dihedral angles and relative surface accessibility of their amino acid residues at the tripeptide level. The proposed approach leverages upon the backbone dihedral angle and surface accessibility preferences of the residues by accounting for its N-terminal and C-terminal neighbors in the protein structure. These preferences are used to evaluate protein structures through a machine learning approach and tested on an extensive dataset of diverse proteins.

RESULTS

The approach was extensively validated on a large test dataset (n = 25 005) of protein structures, comprising 23 661 models of 82 non-homologous proteins and 1344 non-homologous experimental structures. In addition, an external dataset of 40 000 models of 200 non-homologous proteins was also used for the validation of the proposed method. Both datasets were further used for benchmarking the proposed method with four different state-of-the-art methods for protein structure quality assessment. In the benchmarking, the proposed method outperformed some state-of-the-art methods in terms of Spearman's and Pearson's correlation coefficients, average GDT-TS loss, sum of z-scores and average absolute difference of predictions over corresponding observed values. The high accuracy of the proposed approach promises a potential use of the sequence and structural features in computational protein design.

AVAILABILITY AND IMPLEMENTATION

http://github.com/KYZ-LSB/ProTerS-FitFun.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

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.

Missense Mutations Modify the Conformational Ensemble of the α-Synuclein Monomer Which Exhibits a Two-Phase Characteristic

α-Synuclein is an intrinsically disordered protein occurring in different conformations and prone to aggregate in β-sheet structures, which are the hallmark of the Parkinson disease. Missense mutations are associated with familial forms of this neuropathy. How these single amino-acid substitutions modify the conformations of wild-type α-synuclein is unclear. Here, using coarse-grained molecular dynamics simulations, we sampled the conformational space of the wild type and mutants (A30P, A53P, and E46K) of α-synuclein monomers for an effective time scale of 29.7 ms. To characterize the structures, we developed an algorithm, CUTABI (CUrvature and Torsion based of Alpha-helix and Beta-sheet Identification), to identify residues in the α-helix and β-sheet from C^α-coordinates. CUTABI was built from the results of the analysis of 14,652 selected protein structures using the Dictionary of Secondary Structure of Proteins (DSSP) algorithm. DSSP results are reproduced with 93% of success for 10 times lower computational cost. A two-dimensional probability density map of α-synuclein as a function of the number of residues in the α-helix and β-sheet is computed for wild-type and mutated proteins from molecular dynamics trajectories. The density of conformational states reveals a two-phase characteristic with a homogeneous phase (state B, β-sheets) and a heterogeneous phase (state HB, mixture of α-helices and β-sheets). The B state represents 40% of the conformations for the wild-type, A30P, and E46K and only 25% for A53T. The density of conformational states of the B state for A53T and A30P mutants differs from the wild-type one. In addition, the mutant A53T has a larger propensity to form helices than the others. These findings indicate that the equilibrium between the different conformations of the α-synuclein monomer is modified by the missense mutations in a subtle way. The α-helix and β-sheet contents are promising order parameters for intrinsically disordered proteins, whereas other structural properties such as average gyration radius, R_g, or probability distribution of R_g cannot discriminate significantly the conformational ensembles of the wild type and mutants. When separated in states B and HB, the distributions of R_g are more significantly different, indicating that global structural parameters alone are insufficient to characterize the conformational ensembles of the α-synuclein monomer.

Comparative evaluation of shape retrieval methods on macromolecular surfaces: an application of computer vision methods in structural bioinformatics

MOTIVATION

The investigation of the structure of biological systems at the molecular level gives insights about their functions and dynamics. Shape and surface of biomolecules are fundamental to molecular recognition events. Characterizing their geometry can lead to more adequate predictions of their interactions. In the present work, we assess the performance of reference shape retrieval methods from the computer vision community on protein shapes.

RESULTS

Shape retrieval methods are efficient in identifying orthologous proteins and tracking large conformational changes. This work illustrates the interest for the protein surface shape as a higher-level representation of the protein structure that (i) abstracts the underlying protein sequence, structure or fold, (ii) allows the use of shape retrieval methods to screen large databases of protein structures to identify surficial homologs and possible interacting partners and (iii) opens an extension of the protein structure-function paradigm toward a protein structure-surface(s)-function paradigm.

AVAILABILITYAND IMPLEMENTATION

All data are available online at http://datasetmachat.drugdesign.fr.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

SCOPe: improvements to the structural classification of proteins - extended database to facilitate variant interpretation and machine learning

The Structural Classification of Proteins—extended (SCOPe, https://scop.berkeley.edu) knowledgebase aims to provide an accurate, detailed, and comprehensive description of the structural and evolutionary relationships amongst the majority of proteins of known structure, along with resources for analyzing the protein structures and their sequences. Structures from the PDB are divided into domains and classified using a combination of manual curation and highly precise automated methods. In the current release of SCOPe, 2.08, we have developed search and display tools for analysis of genetic variants we mapped to structures classified in SCOPe. In order to improve the utility of SCOPe to automated methods such as deep learning classifiers that rely on multiple alignment of sequences of homologous proteins, we have introduced new machine-parseable annotations that indicate aberrant structures as well as domains that are distinguished by a smaller repeat unit. We also classified structures from 74 of the largest Pfam families not previously classified in SCOPe, and we improved our algorithm to remove N- and C-terminal cloning, expression and purification sequences from SCOPe domains. SCOPe 2.08-stable classifies 106 976 PDB entries (about 60% of PDB entries).

Multi contact-based folding method for de novo protein structure prediction

Meta contact, which combines different contact maps into one to improve contact prediction accuracy and effectively reduce the noise from a single contact map, is a widely used method. However, protein structure prediction using meta contact cannot fully exploit the information carried by original contact maps. In this work, a multi contact-based folding method under the evolutionary algorithm framework, MultiCFold, is proposed. In MultiCFold, the thorough information of different contact maps is directly used by populations to guide protein structure folding. In addition, noncontact is considered as an effective supplement to contact information and can further assist protein folding. MultiCFold is tested on a set of 120 nonredundant proteins, and the average TM-score and average RMSD reach 0.617 and 5.815 Å, respectively. Compared with the meta contact-based method, MetaCFold, average TM-score and average RMSD have a 6.62 and 8.82% improvement. In particular, the import of noncontact information increases the average TM-score by 6.30%. Furthermore, MultiCFold is compared with four state-of-the-art methods of CASP13 on the 24 FM targets, and results show that MultiCFold is significantly better than other methods after the full-atom relax procedure.

Challenging Reverse Screening: A Benchmark Study for Comprehensive Evaluation

As an efficient way of computational target prediction, reverse docking can find not only potential targets but also binding modes for a query ligand. Though the number of available docking tools keeps expanding, there is still not a comprehensive evaluation study which can uncover the advantages and limitations of these strategies in the research field of computational target-fishing. In this study, we propose a brand-new evaluation dataset tailor-made for reverse docking, which is composed of a true positive set (the core set) and two negative sets (the similar decoy set and the dissimilar decoy set). The proposed evaluation dataset can assess the prediction performance of docking tools as various values affected by varying degrees of inter-target ranking bias. The performance of four classical docking programs (AutoDock, AutoDock Vina, Glide and GOLD) was evaluated utilizing our dataset, and a biased prediction performance was observed regarding binding site properties. The results demonstrated that Glide (SP) and Glide(XP) had the best capacity to find true targets whether there was inter-target ranking bias or not.