Protein Sequence Families Research Articles

AlignScape, displaying sequence similarity using self-organizing maps.

The current richness of sequence data needs efficient methodologies to display and analyze the complexity of the information in a compact and readable manner. Traditionally, phylogenetic trees and sequence similarity networks have been used to display and analyze sequences of protein families. These methods aim to shed light on key computational biology problems such as sequence classification and functional inference. Here, we present a new methodology, AlignScape, based on self-organizing maps. AlignScape is applied to three large families of proteins: the kinases and GPCRs from human, and bacterial T6SS proteins. AlignScape provides a map of the similarity landscape and a tree representation of multiple sequence alignments These representations are useful to display, cluster, and classify sequences as well as identify functional trends. The efficient GPU implementation of AlignScape allows the analysis of large MSAs in a few minutes. Furthermore, we show how the AlignScape analysis of proteins belonging to the T6SS complex can be used to predict coevolving partners.

GENERALIST: A latent space based generative model for protein sequence families.

Generative models of protein sequence families are an important tool in the repertoire of protein scientists and engineers alike. However, state-of-the-art generative approaches face inference, accuracy, and overfitting- related obstacles when modeling moderately sized to large proteins and/or protein families with low sequence coverage. Here, we present a simple to learn, tunable, and accurate generative model, GENERALIST: GENERAtive nonLInear tenSor-factorizaTion for protein sequences. GENERALIST accurately captures several high order summary statistics of amino acid covariation. GENERALIST also predicts conservative local optimal sequences which are likely to fold in stable 3D structure. Importantly, unlike current methods, the density of sequences in GENERALIST-modeled sequence ensembles closely resembles the corresponding natural ensembles. Finally, GENERALIST embeds protein sequences in an informative latent space. GENERALIST will be an important tool to study protein sequence variability.

ProtWave-VAE: Integrating Autoregressive Sampling with Latent-Based Inference for Data-Driven Protein Design.

Deep generative models (DGMs) have shown great success in the understanding and data-driven design of proteins. Variational autoencoders (VAEs) are a popular DGM approach that can learn the correlated patterns of amino acid mutations within a multiple sequence alignment (MSA) of protein sequences and distill this information into a low-dimensional latent space to expose phylogenetic and functional relationships and guide generative protein design. Autoregressive (AR) models are another popular DGM approach that typically lacks a low-dimensional latent embedding but does not require training sequences to be aligned into an MSA and enable the design of variable length proteins. In this work, we propose ProtWave-VAE as a novel and lightweight DGM, employing an information maximizing VAE with a dilated convolution encoder and an autoregressive WaveNet decoder. This architecture blends the strengths of the VAE and AR paradigms in enabling training over unaligned sequence data and the conditional generative design of variable length sequences from an interpretable, low-dimensional learned latent space. We evaluated the model's ability to infer patterns and design rules within alignment-free homologous protein family sequences and to design novel synthetic proteins in four diverse protein families. We show that our model can infer meaningful functional and phylogenetic embeddings within latent spaces and make highly accurate predictions within semisupervised downstream fitness prediction tasks. In an application to the C-terminal SH3 domain in the Sho1 transmembrane osmosensing receptor in baker's yeast, we subject ProtWave-VAE-designed sequences to experimental gene synthesis and select-seq assays for the osmosensing function to show that the model enables synthetic protein design, conditional C-terminus diversification, and engineering of the osmosensing function into SH3 paralogues.

An objective criterion to evaluate sequence-similarity networks helps in dividing the protein family sequence space.

The deluge of genomic data raises various challenges for computational protein annotation. The definition of superfamilies, based on conserved folds, or of families, showing more recent homology signatures, allow a first categorization of the sequence space. However, for precise functional annotation or the identification of the unexplored parts within a family, a division into subfamilies is essential. As curators of an expert database, the Carbohydrate Active Enzymes database (CAZy), we began, more than 15 years ago, to manually define subfamilies based on phylogeny reconstruction. However, facing the increasing amount of sequence and functional data, we required more scalable and reproducible methods. The recently popularized sequence similarity networks (SSNs), allows to cope with very large families and computation of many subfamily schemes. Still, the choice of the optimal SSN subfamily scheme only relies on expert knowledge so far, without any data-driven guidance from within the network. In this study, we therefore decided to investigate several network properties to determine a criterion which can be used by curators to evaluate the quality of subfamily assignments. The performance of the closeness centrality criterion, a network property to indicate the connectedness within the network, shows high similarity to the decisions of expert curators from eight distinct protein families. Closeness centrality also suggests that in some cases multiple levels of subfamilies could be possible, depending on the granularity of the research question, while it indicates when no subfamily emerged in some family evolution. We finally used closeness centrality to create subfamilies in four families of the CAZy database, providing a finer functional annotation and highlighting subfamilies without biochemically characterized members for potential future discoveries.

Multi-omics Investigation of Freeze Tolerance in the Amur Sleeper, an Aquatic Ectothermic Vertebrate.

Freeze tolerance, the ability of an organism to survive internal ice formation, is a striking survival strategy employed by some ectotherms living in cold environments. However, the genetic bases of this remarkable adaptation are largely unknown. The Amur sleeper (Perccottus glenii), the only known freeze-tolerant fish species, can overwinter with its entire body frozen in ice. Here, we sequenced the chromosome-level genome of the Amur sleeper and performed comparative genomic, transcriptomic, and metabolomic analyses to investigate its strategies for surviving freezing. Evolutionary analysis suggested that the Amur sleeper diverged from its closest non-cold-hardy relative about 15.07 million years ago and has experienced a high rate of protein evolution. Transcriptomic and metabolomic data identified a coordinated and tissue-specific regulation of genes and metabolites involved in hypometabolism, cellular stress response, and cryoprotectant accumulation involved in freezing and thawing. Several genes show evidence of accelerated protein sequence evolution or family size expansion were found as adaptive responses to freezing-induced stresses. Specifically, genetic changes associated with cytoskeleton stability, cryoprotectant synthesis, transmembrane transport, and neuroprotective adaptations were identified as potentially key innovations that aid in freezing survival. Our work provides valuable resources and opportunities to unveil the molecular adaptations supporting freeze tolerance in ectothermic vertebrates.

Connecting genomic islands across prokaryotic and phage genomes via protein families

Prokaryotic genomes evolve via horizontal gene transfer (HGT), mutations, and rearrangements. A noteworthy part of the HGT process is facilitated by genomic islands (GIs). While previous computational biology research has focused on developing tools to detect GIs in prokaryotic genomes, there has been little research investigating GI patterns and biological connections across species. We have pursued the novel idea of connecting GIs across prokaryotic and phage genomes via patterns of protein families. Such patterns are sequences of protein families frequently present in the genomes of multiple species. We combined the large data set from the IslandViewer4 database with protein families from Pfam while implementing a comprehensive strategy to identify patterns making use of HMMER, BLAST, and MUSCLE. we also implemented Python programs that link the analysis into a single pipeline. Research results demonstrated that related GIs often exist in species that are evolutionarily unrelated and in multiple bacterial phyla. Analysis of the discovered patterns led to the identification of biological connections among prokaryotes and phages. These connections suggest broad HGT connections across the bacterial kingdom and its associated phages. The discovered patterns and connections could provide the basis for additional analysis on HGT breadth and the patterns in pathogenic GIs.

Accuracy of multiple sequence alignment methods in the reconstruction of transposable element families.

The construction of a high-quality multiple sequence alignment (MSA) from copies of a transposable element (TE) is a critical step in the characterization of a new TE family. Most studies of MSA accuracy have been conducted on protein or RNA sequence families, where structural features and strong signals of selection may assist with alignment. Less attention has been given to the quality of sequence alignments involving neutrally evolving DNA sequences such as those resulting from TE replication. Transposable element sequences are challenging to align due to their wide divergence ranges, fragmentation, and predominantly-neutral mutation patterns. To gain insight into the effects of these properties on MSA accuracy, we developed a simulator of TE sequence evolution, and used it to generate a benchmark with which we evaluated the MSA predictions produced by several popular aligners, along with Refiner, a method we developed in the context of our RepeatModeler software. We find that MAFFT and Refiner generally outperform other aligners for low to medium divergence simulated sequences, while Refiner is uniquely effective when tasked with aligning high-divergent and fragmented instances of a family.

Folding the unfoldable: using AlphaFold to explore spurious proteins.

The release of AlphaFold 2.0 has revolutionized our ability to determine protein structures from sequences. This tool also inadvertently opens up many unanticipated opportunities. In this article, we investigate the AntiFam resource, which contains 250 protein sequence families that we believe to be spurious protein translations. We would not expect proteins belonging to these families to fold into well-ordered globular structures. To test this hypothesis, we have attempted to computationally determine the structure of a representative sequence from all AntiFam 6.0 families. Although the large majority of families showed no evidence of globular structure, we have identified one example for which a globular structure is predicted. Proteins in this AntiFam entry indeed seem likely to be bona fide proteins, based on additional considerations, and thus AlphaFold provides a useful quality control for the AntiFam database. Conversely, known spurious proteins offer useful set of quality controls for AlphaFold. We have identified a trend that the mean structure prediction confidence score pLDDT is higher for shorter sequences. Of the 131 AntiFam representative sequences <100 amino acids in length, AlphaFold predicts a mean pLDDT of 80 or greater for six of them. Thus, particular care should be taken when applying AlphaFold to short protein sequences. The AlphaFold predictions for representative sequences can be found at the following URL: https://drive.google.com/drive/folders/1u9OocRIAabGQn56GljoG1JTDAxjkY1ro. Supplementary data are available at Bioinformatics Advances online.

TIAMMAt: Leveraging Biodiversity to Revise Protein Domain Models, Evidence from Innate Immunity.

Sequence annotation is fundamental for studying the evolution of protein families, particularly when working with nonmodel species. Given the rapid, ever-increasing number of species receiving high-quality genome sequencing, accurate domain modeling that is representative of species diversity is crucial for understanding protein family sequence evolution and their inferred function(s). Here, we describe a bioinformatic tool called Taxon-Informed Adjustment of Markov Model Attributes (TIAMMAt) which revises domain profile hidden Markov models (HMMs) by incorporating homologous domain sequences from underrepresented and nonmodel species. Using innate immunity pathways as a case study, we show that revising profile HMM parameters to directly account for variation in homologs among underrepresented species provides valuable insight into the evolution of protein families. Following adjustment by TIAMMAt, domain profile HMMs exhibit changes in their per-site amino acid state emission probabilities and insertion/deletion probabilities while maintaining the overall structure of the consensus sequence. Our results show that domain revision can heavily impact evolutionary interpretations for some families (i.e., NLR’s NACHT domain), whereas impact on other domains (e.g., rel homology domain and interferon regulatory factor domains) is minimal due to high levels of sequence conservation across the sampled phylogenetic depth (i.e., Metazoa). Importantly, TIAMMAt revises target domain models to reflect homologous sequence variation using the taxonomic distribution under consideration by the user. TIAMMAt’s flexibility to revise any subset of the Pfam database using a user-defined taxonomic pool will make it a valuable tool for future protein evolution studies, particularly when incorporating (or focusing) on nonmodel species.

Improving sequence-based modeling of protein families using secondary-structure quality assessment.

MotivationModeling of protein family sequence distribution from homologous sequence data recently received considerable attention, in particular for structure and function predictions, as well as for protein design. In particular, direct coupling analysis, a method to infer effective pairwise interactions between residues, was shown to capture important structural constraints and to successfully generate functional protein sequences. Building on this and other graphical models, we introduce a new framework to assess the quality of the secondary structures of the generated sequences with respect to reference structures for the family.ResultsWe introduce two scoring functions characterizing the likeliness of the secondary structure of a protein sequence to match a reference structure, called Dot Product and Pattern Matching. We test these scores on published experimental protein mutagenesis and design dataset, and show improvement in the detection of nonfunctional sequences. We also show that use of these scores help rejecting nonfunctional sequences generated by graphical models (Restricted Boltzmann Machines) learned from homologous sequence alignments.Availability and implementationData and code available at https://github.com/CyrilMa/ssqaSupplementary information Supplementary data are available at Bioinformatics online.

TSFM 1.0: tRNA Structure-Function Mapper.

Structure-conditioned information statistics have proven useful to predict and visualize tRNA Class-Informative Features (CIFs) and their evolutionary divergences. Although permutation P-values can quantify the significance of CIF divergences between two taxa, their naive Monte Carlo approximation is slow and inaccurate. The Peaks-over-Threshold approach of Knijnenburg et al. (2009) promises improvements to both speed and accuracy of permutation P-values, but has no publicly available API. We present tRNA Structure-Function Mapper (tSFM) v1.0, an open-source, multi-threaded application that efficiently computes, visualizes and assesses significance of single- and paired-site CIFs and their evolutionary divergences for any RNA, protein, gene or genomic element sequence family. Multiple estimators of permutation P-values for CIF evolutionary divergences are provided along with confidence intervals. tSFM is implemented in Python 3 with compiled C extensions and is freely available through GitHub (https://github.com/tlawrence3/tSFM) and PyPI. The data underlying this article are available on GitHub at https://github.com/tlawrence3/tSFM. Supplementary data are available at Bioinformatics online.

DeepPPF: A deep learning framework for predicting protein family

Machine learning pipelines for protein functional family prediction are urgently needed especially now that only 1% of raw protein sequences have been manually annotated. Although existing machine learning algorithms have achieved a decent performance in modeling and predicting the functional families of protein sequences, they still have two drawbacks. First, biological dependencies among nucleotides are not rich enough to describe motifs for these methods. Also, existing algorithms are not accurate enough to predict the functional families of newly discovered proteins. To address the above limitations simultaneously, we propose a novel deep learning framework for predicting protein family, DeepPPF, which employs the word2vec technique in capturing distributional dependencies among nucleotides and discovers rich features from diverse motif lengths to characterize proteins. The novelty of the DeepPPF is in utilizing distributional dependencies among nucleotides. Experimental results on G protein-coupled receptor hierarchical datasets show the effectiveness of DeepPPF in achieving the state-of-the-art performance in items of Mathew’s correlation coefficients (MCC) of 97.62%, 88.45% and, 83.09% for family, sub-family and, sub-subfamily hierarchical levels, respectively. Also, DeepPPF outperformed existing methods in terms of prediction accuracy and Mathew’s correlation coefficients on the cluster of orthologous groups (COG) and phage of orthologous groups (POG) datasets. Furthermore, we analyzed the ability of DeepPPF framework to discover rich motifs for functional classes with the least sets of protein sequences. The experimental results show that rich motif discovery is key to improving the modeling performance of protein families through deep learning techniques. Finally, we investigated the effect of transferring a low-level functional domain level to a high-level functional domain and results show that the target domain prediction can be improved with transfer learning. Therefore, our proposed deep learning framework can be useful in characterizing protein functional families. The codes and datasets are available at https://github.com/CSUBioGroup/DeepPPF.

Gains and losses of metabolic function inferred from a phylotranscriptomic analysis of algae

Hidden Markov models representing 167 protein sequence families were used to infer the presence or absence of homologs within the transcriptomes of 183 algal species/strains. Statistical analyses of the distribution of HMM hits across major clades of algae, or at branch points on the phylogenetic tree of 98 chlorophytes, confirmed and extended known cases of metabolic loss and gain, most notably the loss of the mevalonate pathway for terpenoid synthesis in green algae but not, as we show here, in the streptophyte algae. Evidence for novel events was found as well, most remarkably in the recurrent and coordinated gain or loss of enzymes for the glyoxylate shunt. We find, as well, a curious pattern of retention (or re-gain) of HMG-CoA synthase in chlorophytes that have otherwise lost the mevalonate pathway, suggesting a novel, co-opted function for this enzyme in select lineages. Finally, we find striking, phylogenetically linked distributions of coding sequences for three pathways that synthesize the major membrane lipid phosphatidylcholine, and a complementary phylogenetic distribution pattern for the non-phospholipid DGTS (diacyl-glyceryl-trimethylhomoserine). Mass spectrometric analysis of lipids from 25 species was used to validate the inference of DGTS synthesis from sequence data.

Estimating probabilistic context-free grammars for proteins using contact map constraints.

Interactions between amino acids that are close in the spatial structure, but not necessarily in the sequence, play important structural and functional roles in proteins. These non-local interactions ought to be taken into account when modeling collections of proteins. Yet the most popular representations of sets of related protein sequences remain the profile Hidden Markov Models. By modeling independently the distributions of the conserved columns from an underlying multiple sequence alignment of the proteins, these models are unable to capture dependencies between the protein residues. Non-local interactions can be represented by using more expressive grammatical models. However, learning such grammars is difficult. In this work, we propose to use information on protein contacts to facilitate the training of probabilistic context-free grammars representing families of protein sequences. We develop the theory behind the introduction of contact constraints in maximum-likelihood and contrastive estimation schemes and implement it in a machine learning framework for protein grammars. The proposed framework is tested on samples of protein motifs in comparison with learning without contact constraints. The evaluation shows high fidelity of grammatical descriptors to protein structures and improved precision in recognizing sequences. Finally, we present an example of using our method in a practical setting and demonstrate its potential beyond the current state of the art by creating a grammatical model of a meta-family of protein motifs. We conclude that the current piece of research is a significant step towards more flexible and accurate modeling of collections of protein sequences. The software package is made available to the community.

Sequence-Based Fingerprinting of Intrinsically Disordered Regions

Sequence conservation and the high fidelity of sequence-to-structure relationships enable the identification of protein sequence families. This is especially true of proteins that are capable of folding as autonomous units. However, over a third of the eukaryotic proteome consists of intrinsically disordered proteins and regions (IDRs) that either fold upon binding into distinct context-dependent structures or persist in disordered albeit functional states. Freed from the constraint of autonomous folding, the sequences of IDRs are frequently less well conserved as assessed by standard alignment-based metrics. Accordingly, traditional methods for sequence classification often fail when identifying sequence families that are defined by IDRs. The apparent lack of sequence conservation could be taken to mean that random sequences are interoperable with the sequences of IDRs, yet several recent studies show that this is clearly not the case. Biophysical studies have helped identify the existence of sequence-to-conformation relationships for IDPs. These relationships, in turn, govern the molecular functions of IDRs and by extrapolation we propose that there is an amino-acid based grammar that underlies sequence-to-function relationships as well. Sequence complexity, amino acid composition, and the linear patterning of different types of amino acids are among the main determinants of sequence-to-conformation relationships and the consequences for functions and phenotypes. We have used these sequence features as fingerprints to develop ways to delineate sequence families for IDRs. These measures are designed to enable the effective classification and comparison of IDRs, with applications towards sequence design, evolutionary analysis, and ultimately the prediction of molecular functions of IDRs. These efforts are likely to enable an improved understanding of the evolutionary determinants of IDRs versus intrinsically foldable regions.

On the Natural Structure of Amino Acid Patterns in Families of Protein Sequences.

All known terrestrial proteins are coded as continuous strings of ≈20 amino acids. The patterns formed by the repetitions of elements in groups of finite sequences describes the natural architectures of protein families. We present a method to search for patterns and groupings of patterns in protein sequences using a mathematically precise definition for "repetition", an efficient algorithmic implementation and a robust scoring system with no adjustable parameters. We show that the sequence patterns can be well-separated into disjoint classes according to their recurrence in nested structures. The statistics of the occurrences of patterns indicate that short repetitions are sufficient to account for the differences between natural families and randomized groups of sequences by more than 10 standard deviations, while contiguous sequence patterns shorter than 5 residues are effectively random in their occurrences. A small subset of patterns is sufficient to account for a robust "familiarity" definition between arbitrary sets of sequences.

Protein structure prediction: making AWSEM AWSEM-ER by adding evolutionary restraints.

Protein sequences have evolved to fold into functional structures, resulting in families of diverse protein sequences that all share the same overall fold. One can harness protein family sequence data to infer likely contacts between pairs of residues. In the current study, we combine this kind of inference from coevolutionary information with a coarse-grained protein force field ordinarily used with single sequence input, the Associative memory, Water mediated, Structure and Energy Model (AWSEM), to achieve improved structure prediction. The resulting Associative memory, Water mediated, Structure and Energy Model with Evolutionary Restraints (AWSEM-ER) yields a significant improvement in the quality of protein structure prediction over the single sequence prediction from AWSEM when a sufficiently large number of homologous sequences are available. Free energy landscape analysis shows that the addition of the evolutionary term shifts the free energy minimum to more native-like structures, which explains the improvement in the quality of structures when performing predictions using simulated annealing. Simulations using AWSEM without coevolutionary information have proved useful in elucidating not only protein folding behavior, but also mechanisms of protein function. The success of AWSEM-ER in de novo structure prediction suggests that the enhanced model opens the door to functional studies of proteins even when no experimentally solved structures are available.

Genomic Enzymology: Web Tools for Leveraging Protein Family Sequence-Function Space and Genome Context to Discover Novel Functions.

The exponentially increasing number of protein and nucleic acid sequences provides opportunities to discover novel enzymes, metabolic pathways, and metabolites/natural products, thereby adding to our knowledge of biochemistry and biology. The challenge has evolved from generating sequence information to mining the databases to integrating and leveraging the available information, i.e., the availability of “genomic enzymology” web tools. Web tools that allow identification of biosynthetic gene clusters are widely used by the natural products/synthetic biology community, thereby facilitating the discovery of novel natural products and the enzymes responsible for their biosynthesis. However, many novel enzymes with interesting mechanisms participate in uncharacterized small-molecule metabolic pathways; their discovery and functional characterization also can be accomplished by leveraging information in protein and nucleic acid databases. This Perspective focuses on two genomic enzymology web tools that assist the discovery novel metabolic pathways: (1) Enzyme Function Initiative-Enzyme Similarity Tool (EFI-EST) for generating sequence similarity networks to visualize and analyze sequence–function space in protein families and (2) Enzyme Function Initiative-Genome Neighborhood Tool (EFI-GNT) for generating genome neighborhood networks to visualize and analyze the genome context in microbial and fungal genomes. Both tools have been adapted to other applications to facilitate target selection for enzyme discovery and functional characterization. As the natural products community has demonstrated, the enzymology community needs to embrace the essential role of web tools that allow the protein and genome sequence databases to be leveraged for novel insights into enzymological problems.

The crystal structure of Erwinia amylovora AmyR, a member of the YbjN protein family, shows similarity to type III secretion chaperones but suggests different cellular functions.

AmyR is a stress and virulence associated protein from the plant pathogenic Enterobacteriaceae species Erwinia amylovora, and is a functionally conserved ortholog of YbjN from Escherichia coli. The crystal structure of E. amylovora AmyR reveals a class I type III secretion chaperone-like fold, despite the lack of sequence similarity between these two classes of protein and lacking any evidence of a secretion-associated role. The results indicate that AmyR, and YbjN proteins in general, function through protein-protein interactions without any enzymatic action. The YbjN proteins of Enterobacteriaceae show remarkably low sequence similarity with other members of the YbjN protein family in Eubacteria, yet a high level of structural conservation is observed. Across the YbjN protein family sequence conservation is limited to residues stabilising the protein core and dimerization interface, while interacting regions are only conserved between closely related species. This study presents the first structure of a YbjN protein from Enterobacteriaceae, the most highly divergent and well-studied subgroup of YbjN proteins, and an in-depth sequence and structural analysis of this important but poorly understood protein family.

Multiple Sequence Alignment.

The increasing importance of Next Generation Sequencing (NGS) techniques has highlighted the key role of multiple sequence alignment (MSA) in comparative structure and function analysis of biological sequences. MSA often leads to fundamental biological insight into sequence-structure-function relationships of nucleotide or protein sequence families. Significant advances have been achieved in this field, and many useful tools have been developed for constructing alignments, although many biological and methodological issues are still open. This chapter first provides some background information and considerations associated with MSA techniques, concentrating on the alignment of protein sequences. Then, a practical overview of currently available methods and a description of their specific advantages and limitations are given, to serve as a helpful guide or starting point for researchers who aim to construct a reliable MSA.