Remote Homology Detection Research Articles

Unique structural features and sequence motifs of proline utilization A (PutA)

Proline utilization A proteins (PutAs) are bifunctional enzymes that catalyze the oxidation of proline to glutamate using spatially separated proline dehydrogenase and pyrroline-5-carboxylate dehydrogenase active sites. Here we use the crystal structure of the minimalist PutA from Bradyrhizobium japonicum (BjPutA) along with sequence analysis to identify unique structural features of PutAs. This analysis shows that PutAs have secondary structural elements and domains not found in the related monofunctional enzymes. Some of these extra features are predicted to be important for substrate channeling in BjPutA. Multiple sequence alignment analysis shows that some PutAs have a 17-residue conserved motif in the C-terminal 20-30 residues of the polypeptide chain. The BjPutA structure shows that this motif helps seal the internal substrate-channeling cavity from the bulk medium. Finally, it is shown that some PutAs have a 100-200 residue domain of unknown function in the C-terminus that is not found in minimalist PutAs. Remote homology detection suggests that this domain is homologous to the oligomerization beta-hairpin and Rossmann fold domain of BjPutA.

Remote homologue identification of Drosophila GAGA factor in mouse.

GAGA factor (GAF) is involved in both gene activation and gene repression and plays a role in the modulation of chromatin structure. In Drosophila, Trithroax like (Trl) gene encodes the DNA binding protein called GAGA factor (GAF). Trl-GAF binds to GAGA sites through its C2H2 zinc finger domain and has an N-terminal BTB/POZ domain. Identification of Trl-GAF homologue in mouse helps in deeper understanding of the mechanism and function. Conventional alignment tools such as BLAST and FASTA cannot identify homologues in mouse genome as their sequence identity is below 30%. In the present study, various sequence and structure analyses were followed for the detection of remote homologues of Drosophila GAGA FACTOR in mouse to identify as Zbtb3. Through homology modeling and docking approach, the zinc finger region of mouse Zbtb3 showed conserved residues and favorable DNA binding sites with GAGA sites similar to that of Drosophila GAGA FACTOR.

AlignHUSH: Alignment of HMMs using structure and hydrophobicity information

BackgroundSensitive remote homology detection and accurate alignments especially in the midnight zone of sequence similarity are needed for better function annotation and structural modeling of proteins. An algorithm, AlignHUSH for HMM-HMM alignment has been developed which is capable of recognizing distantly related domain families The method uses structural information, in the form of predicted secondary structure probabilities, and hydrophobicity of amino acids to align HMMs of two sets of aligned sequences. The effect of using adjoining column(s) information has also been investigated and is found to increase the sensitivity of HMM-HMM alignments and remote homology detection.ResultsWe have assessed the performance of AlignHUSH using known evolutionary relationships available in SCOP. AlignHUSH performs better than the best HMM-HMM alignment methods and is observed to be even more sensitive at higher error rates. Accuracy of the alignments obtained using AlignHUSH has been assessed using the structure-based alignments available in BaliBASE. The alignment length and the alignment quality are found to be appropriate for homology modeling and function annotation. The alignment accuracy is found to be comparable to existing methods for profile-profile alignments.ConclusionsA new method to align HMMs has been developed and is shown to have better sensitivity at error rates of 10% and above when compared to other available programs. The proposed method could effectively aid obtaining clues to functions of proteins of yet unknown function.A web-server incorporating the AlignHUSH method is available at http://crick.mbu.iisc.ernet.in/~alignhush/

Remote protein homology detection and fold recognition using two-layer support vector machine classifiers

Remote protein homology detection and fold recognition refer to detection of structural homology in proteins where there are small or no similarities in the sequence. To detect protein structural classes from protein primary sequence information, homology-based methods have been developed, which can be divided to three types: discriminative classifiers, generative models for protein families and pairwise sequence comparisons. Support Vector Machines (SVM) and Neural Networks (NN) are two popular discriminative methods. Recent studies have shown that SVM has fast speed during training, more accurate and efficient compared to NN. We present a comprehensive method based on two-layer classifiers. The 1st layer is used to detect up to superfamily and family in SCOP hierarchy using optimized binary SVM classification rules. It used the kernel function known as the Bio-kernel, which incorporates the biological information in the classification process. The 2nd layer uses discriminative SVM algorithm with string kernel that will detect up to protein fold level in SCOP hierarchy. The results obtained were evaluated using mean ROC and mean MRFP and the significance of the result produced with pairwise t-test was tested. Experimental results show that our approaches significantly improve the performance of remote protein homology detection and fold recognition for all three different version SCOP datasets (1.53, 1.67 and 1.73). We achieved 4.19% improvements in term of mean ROC in SCOP 1.53, 4.75% in SCOP 1.67 and 4.03% in SCOP 1.73 datasets when compared to the result produced by well-known methods. The combination of first layer and second layer of BioSVM-2L performs well in remote homology detection and fold recognition even in three different versions of datasets.

Protein remote homology detection based on auto-cross covariance transformation

Protein remote homology detection is a critical step toward annotating its structure and function. Supervised learning algorithms such as support vector machine are currently the most accurate methods. The position-specific score matrices (PSSMs) contain wealthy information about the evolutionary relationship of proteins. However, the PSSMs often have different lengths, which are difficult to be used by machine-learning methods. In this study, a simple, fast and powerful method is presented for protein remote homology detection, which combines support vector machine with auto-cross covariance transformation. The PSSMs are converted into a series of fixed-length vectors by auto-cross covariance transformation and these vectors are then input to a support vector machine classifier for remote homology detection. The sequence-order effects can be effectively captured by this scheme. Experiments are performed on well-established datasets, and the remote homology is simulated at the superfamily and the fold level, respectively. The results show that the proposed method, referred to as ACCRe, is comparable or even better than the state-of-the-art methods in terms of detection performance, and its time complexity is superior to those of other profile-based SVM methods. The auto-cross covariance transformation provides a novel way for the usage of evolutionary information, which can be widely used for protein-level studies.

The Irredundant Class Method for Remote Homology Detection of Protein Sequences

The automatic classification of protein sequences into families is of great help for the functional prediction and annotation of new proteins. In this article, we present a method called Irredundant Class that address the remote homology detection problem. The best performing methods that solve this problem are string kernels, that compute a similarity function between pairs of proteins based on their subsequence composition. We provide evidence that almost all string kernels are based on patterns that are not independent, and therefore the associated similarity scores are obtained using a set of redundant features, overestimating the similarity between the proteins. To specifically address this issue, we introduce the class of irredundant common patterns. Loosely speaking, the set of irredundant common patterns is the smallest class of independent patterns that can describe all common patterns in a pair of sequences. We present a classification method based on the statistics of these patterns, named Irredundant Class. Results on benchmark data show that the Irredundant Class outperforms most of the string kernels previously proposed, and it achieves results as good as the current state-of-the-art method Local Alignment, but using the same pairwise information only once.

Remote Thioredoxin Recognition Using Evolutionary Conservation and Structural Dynamics

The thioredoxin family of oxidoreductases plays an important role in redox signaling and control of protein function. Not only are thioredoxins linked to a variety of disorders, but their stable structure has also seen application in protein engineering. Both sequence-based and structure-based tools exist for thioredoxin identification, but remote homolog detection remains a challenge. We developed a thioredoxin predictor using the approach of integrating sequence with structural information. We combined a sequence-based Hidden Markov Model (HMM) with a molecular dynamics enhanced structure-based recognition method (dynamic FEATURE, DF). This hybrid method (HMMDF) has high precision and recall (0.90 and 0.95, respectively) compared with HMM (0.92 and 0.87, respectively) and DF (0.82 and 0.97, respectively). Dynamic FEATURE is sensitive but struggles to resolve closely related protein families, while HMM identifies these evolutionary differences by compromising sensitivity. Our method applied to structural genomics targets makesa strong prediction of a novel thioredoxin.

A discriminative method for family-based protein remote homology detection that combines inductive logic programming and propositional models.

BackgroundRemote homology detection is a hard computational problem. Most approaches have trained computational models by using either full protein sequences or multiple sequence alignments (MSA), including all positions. However, when we deal with proteins in the "twilight zone" we can observe that only some segments of sequences (motifs) are conserved. We introduce a novel logical representation that allows us to represent physico-chemical properties of sequences, conserved amino acid positions and conserved physico-chemical positions in the MSA. From this, Inductive Logic Programming (ILP) finds the most frequent patterns (motifs) and uses them to train propositional models, such as decision trees and support vector machines (SVM).ResultsWe use the SCOP database to perform our experiments by evaluating protein recognition within the same superfamily. Our results show that our methodology when using SVM performs significantly better than some of the state of the art methods, and comparable to other. However, our method provides a comprehensible set of logical rules that can help to understand what determines a protein function.ConclusionsThe strategy of selecting only the most frequent patterns is effective for the remote homology detection. This is possible through a suitable first-order logical representation of homologous properties, and through a set of frequent patterns, found by an ILP system, that summarizes essential features of protein functions.

Detecting remote evolutionary relationships among proteins by large-scale semantic embedding.

Virtually every molecular biologist has searched a protein or DNA sequence database to find sequences that are evolutionarily related to a given query. Pairwise sequence comparison methods—i.e., measures of similarity between query and target sequences—provide the engine for sequence database search and have been the subject of 30 years of computational research. For the difficult problem of detecting remote evolutionary relationships between protein sequences, the most successful pairwise comparison methods involve building local models (e.g., profile hidden Markov models) of protein sequences. However, recent work in massive data domains like web search and natural language processing demonstrate the advantage of exploiting the global structure of the data space. Motivated by this work, we present a large-scale algorithm called ProtEmbed, which learns an embedding of protein sequences into a low-dimensional “semantic space.” Evolutionarily related proteins are embedded in close proximity, and additional pieces of evidence, such as 3D structural similarity or class labels, can be incorporated into the learning process. We find that ProtEmbed achieves superior accuracy to widely used pairwise sequence methods like PSI-BLAST and HHSearch for remote homology detection; it also outperforms our previous RankProp algorithm, which incorporates global structure in the form of a protein similarity network. Finally, the ProtEmbed embedding space can be visualized, both at the global level and local to a given query, yielding intuition about the structure of protein sequence space.

Protein Remote Homology Detection and Fold Recognition based on Features Extracted from Frequency Profiles

Protein remote homology detection and fold recognition are central problems in bioinformatics. Currently, discriminative methods based on support vector machine (SVM) are the most effective and accurate methods for solving these problems. The performance of SVM depends on the method of protein vectorization, so a suitable representation of the protein sequence is a key step for the SVM-based methods. In this paper, two kinds of profile-level building blocks of proteins, binary profiles and N-nary profiles, have been presented, which contain the evolutionary information of the protein sequence frequency profile. The protein sequence frequency profiles calculated from the multiple sequence alignments outputted by PSI-BLAST are converted into binary profiles or N-nary profiles. The protein sequences are transformed into fixed-dimension feature vectors by the occurrence times of each binary profile or N-nary profile and then the corresponding vectors are inputted to support vector machines. The latent semantic analysis (LSA) model, an efficient feature extraction algorithm, is adopted to further improve the performance of our methods. Experiments with protein remote homology detection and fold recognition show that the methods based on profile-level building blocks give better results compared to related methods.

Influence of Genomic and Other Biological Data Sets in the Understanding of Protein Structures, Functions and Interactions

In the post-genomic era, biological databases are growing at a tremendous rate. Despite rapid accumulation of biological information, functions and other biological properties of many putative gene products of various organisms remain either unknown or obscure. This paper examines how strategic integration of large biological databases and combinations of various biological information helps address some of the fundamental questions on protein structure, function and interactions. New developments in function recognition by remote homology detection and strategic use of sequence databases aid recognition of functions of newly discovered proteins. Knowledge of 3-D structures and combined use of sequences and 3-D structures of homologous protein domains expands the ability of remote homology detection enormously. The authors also demonstrate how combined consideration of functions of individual domains of multi-domain proteins helps in recognizing gross biological attributes. This paper also discusses a few cases of combining disparate biological datasets or combination of disparate biological information in obtaining new insights about protein-protein interactions across a host and a pathogen. Finally, the authors discuss how combinations of low resolution structural data, obtained using cryoEM studies, of gigantic multi-component assemblies, and atomic level 3-D structures of the components is effective in inferring finer features in the assembly.

Of bits and bugs--on the use of bioinformatics and a bacterial crystal structure to solve a eukaryotic repeat-protein structure.

Pur-α is a nucleic acid-binding protein involved in cell cycle control, transcription, and neuronal function. Initially no prediction of the three-dimensional structure of Pur-α was possible. However, recently we solved the X-ray structure of Pur-α from the fruitfly Drosophila melanogaster and showed that it contains a so-called PUR domain. Here we explain how we exploited bioinformatics tools in combination with X-ray structure determination of a bacterial homolog to obtain diffracting crystals and the high-resolution structure of Drosophila Pur-α. First, we used sensitive methods for remote-homology detection to find three repetitive regions in Pur-α. We realized that our lack of understanding how these repeats interact to form a globular domain was a major problem for crystallization and structure determination. With our information on the repeat motifs we then identified a distant bacterial homolog that contains only one repeat. We determined the bacterial crystal structure and found that two of the repeats interact to form a globular domain. Based on this bacterial structure, we calculated a computational model of the eukaryotic protein. The model allowed us to design a crystallizable fragment and to determine the structure of Drosophila Pur-α. Key for success was the fact that single repeats of the bacterial protein self-assembled into a globular domain, instructing us on the number and boundaries of repeats to be included for crystallization trials with the eukaryotic protein. This study demonstrates that the simpler structural domain arrangement of a distant prokaryotic protein can guide the design of eukaryotic crystallization constructs. Since many eukaryotic proteins contain multiple repeats or repeating domains, this approach might be instructive for structural studies of a range of proteins.

Hidden Markov model speed heuristic and iterative HMM search procedure

BackgroundProfile hidden Markov models (profile-HMMs) are sensitive tools for remote protein homology detection, but the main scoring algorithms, Viterbi or Forward, require considerable time to search large sequence databases.ResultsWe have designed a series of database filtering steps, HMMERHEAD, that are applied prior to the scoring algorithms, as implemented in the HMMER package, in an effort to reduce search time. Using this heuristic, we obtain a 20-fold decrease in Forward and a 6-fold decrease in Viterbi search time with a minimal loss in sensitivity relative to the unfiltered approaches. We then implemented an iterative profile-HMM search method, JackHMMER, which employs the HMMERHEAD heuristic. Due to our search heuristic, we eliminated the subdatabase creation that is common in current iterative profile-HMM approaches. On our benchmark, JackHMMER detects 14% more remote protein homologs than SAM's iterative method T2K.ConclusionsOur search heuristic, HMMERHEAD, significantly reduces the time needed to score a profile-HMM against large sequence databases. This search heuristic allowed us to implement an iterative profile-HMM search method, JackHMMER, which detects significantly more remote protein homologs than SAM's T2K and NCBI's PSI-BLAST.

Fold homology detection using sequence fragment composition profiles of proteins

The effectiveness of sequence alignment in detecting structural homology among protein sequences decreases markedly when pairwise sequence identity is low (the so-called "twilight zone" problem of sequence alignment). Alternative sequence comparison strategies able to detect structural kinship among highly divergent sequences are necessary to address this need. Among them are alignment-free methods, which use global sequence properties (such as amino acid composition) to identify structural homology in a rapid and straightforward way. We explore the viability of using tetramer sequence fragment composition profiles in finding structural relationships that lie undetected by traditional alignment. We establish a strategy to recast any given protein sequence into a tetramer sequence fragment composition profile, using a series of amino acid clustering steps that have been optimized for mutual information. Our method has the effect of compressing the set of 160,000 unique tetramers (if using the 20-letter amino acid alphabet) into a more tractable number of reduced tetramers (approximately 15-30), so that a meaningful tetramer composition profile can be constructed. We test remote homology detection at the topology and fold superfamily levels using a comprehensive set of fold homologs, culled from the CATH database that share low pairwise sequence similarity. Using the receiver-operating characteristic measure, we demonstrate potentially significant improvement in using information-optimized reduced tetramer composition, over methods relying only on the raw amino acid composition or on traditional sequence alignment, in homology detection at or below the "twilight zone".

Recognition of beta-structural motifs using hidden Markov models trained with simulated evolution

Motivation: One of the most successful methods to date for recognizing protein sequences that are evolutionarily related, has been profile hidden Markov models. However, these models do not capture pairwise statistical preferences of residues that are hydrogen bonded in β-sheets. We thus explore methods for incorporating pairwise dependencies into these models.Results: We consider the remote homology detection problem for β-structural motifs. In particular, we ask if a statistical model trained on members of only one family in a SCOP β-structural superfamily, can recognize members of other families in that superfamily. We show that HMMs trained with our pairwise model of simulated evolution achieve nearly a median 5% improvement in AUC for β-structural motif recognition as compared to ordinary HMMs.Availability: All datasets and HMMs are available at: http://bcb.cs.tufts.edu/pairwise/Contact: anoop.kumar@tufts.edu; lenore.cowen@tufts.edu

Physicochemical property distributions for accurate and rapid pairwise protein homology detection

BackgroundThe challenge of remote homology detection is that many evolutionarily related sequences have very little similarity at the amino acid level. Kernel-based discriminative methods, such as support vector machines (SVMs), that use vector representations of sequences derived from sequence properties have been shown to have superior accuracy when compared to traditional approaches for the task of remote homology detection.ResultsWe introduce a new method for feature vector representation based on the physicochemical properties of the primary protein sequence. A distribution of physicochemical property scores are assembled from 4-mers of the sequence and normalized based on the null distribution of the property over all possible 4-mers. With this approach there is little computational cost associated with the transformation of the protein into feature space, and overall performance in terms of remote homology detection is comparable with current state-of-the-art methods. We demonstrate that the features can be used for the task of pairwise remote homology detection with improved accuracy versus sequence-based methods such as BLAST and other feature-based methods of similar computational cost.ConclusionsA protein feature method based on physicochemical properties is a viable approach for extracting features in a computationally inexpensive manner while retaining the sensitivity of SVM protein homology detection. Furthermore, identifying features that can be used for generic pairwise homology detection in lieu of family-based homology detection is important for applications such as large database searches and comparative genomics.

On the origin of MADS-domain transcription factors

MADS-domain transcription factors are involved in signal transduction and developmental control in plants, animals and fungi. Because their diversification is linked to the origin of novelties in multicellular eukaryotes, the early evolution of MADS-domain proteins is of interest, but has remained enigmatic. Employing whole genome sequence information and remote homology detection methods, we demonstrate that the MADS domain originated from a region of topoisomerases IIA subunit A. Furthermore, we provide evidence that gene duplication occurred in the lineage that led to the MRCA of extant eukaryotes, giving rise to SRF-like and MEF2-like MADS-box genes.

Genomic assessment of the evolution of the prion protein gene family in vertebrates

Prion diseases are devastating neurological disorders caused by the propagation of particles containing an alternative β−sheet-rich form of the prion protein (PrP). Genes paralogous to PrP, called Doppel and Shadoo, have been identified, that also have neuropathological relevance. To aid in the further functional characterization of PrP and its relatives, we annotated completely the PrP gene family (PrP-GF), in the genomes of 42 vertebrates, through combined strategic application of gene prediction programs and advanced remote homology detection techniques (such as HMMs, PSI-TBLASTN and pGenThreader). We have uncovered several previously undescribed paralogous genes and pseudogenes. We find that current high-quality genomic evidence indicates that the PrP relative Doppel, was likely present in the last common ancestor of present-day Tetrapoda, but was lost in the bird lineage, since its divergence from reptiles. Using the new gene annotations, we have defined the consensus of structural features that are characteristic of the PrP and Doppel structures, across diverse Tetrapoda clades. Furthermore, we describe in detail a transcribed pseudogene derived from Shadoo that is conserved across primates, and that overlaps the meiosis gene, SYCE1, thus possibly regulating its expression. In addition, we analysed the locus of PRNP/ PRND for significant conservation across the genomic DNA of eleven mammals, and determined the phylogenetic penetration of non-coding exons. The genomic evidence indicates that the second PRNP non-coding exon found in even-toed ungulates and rodents, is conserved in all high-coverage genome assemblies of primates (human, chimp, orang utan and macaque), and is, at least, likely to have fallen out of use during primate speciation. Furthermore, we have demonstrated that the PRNT gene (at the PRNP human locus) is conserved across at least sixteen mammals, and evolves like a long non-coding RNA, fashioned from fragments of ancient, long, interspersed elements. These annotations and evolutionary analyses will be of further use for functional characterisation of the PrP-GF, and will be updatable in a semi-automated fashion as more genomes accumulate.

Exploring genomes for glycosyltransferases

Glycosyltransferases are one of the largest and most diverse enzyme groups in Nature. They catalyse the synthesis of glycosidic linkages by the transfer of a sugar residue from a donor to an acceptor substrate. These enzymes have been classified into families on the basis of amino acid sequence similarity that are kept updated in the Carbohydrate Active enZyme database (CAZy, ). The repertoire of glycosyltransferases in genomes is believed to determine the diversity of cellular glycan structures, and current estimates suggest that for most genomes about 1% of the coding regions are glycosyltransferases. However, plants tend to have far more glycosyltransferase genes than any other organism sequenced to date, and this can be explained by the highly complex polysaccharide network that form the cell wall and also by the numerous glycosylated secondary metabolites. In recent years, various bioinformatics strategies have been used to search bacterial and plant genomes for new glycosyltransferase genes. These are based on the use of remote homology detection methods that act at the 1D, 2D, and 3D level. The combined use of methods such as profile Hidden Markov Model (HMM) and fold recognition appears to be appropriate for this class of enzyme. Chemometric tools are also particularly well suited for obtaining an overview of multivariate data and revealing hidden latent information when dealing with large and highly complex datasets.

3PFDB--a database of best representative PSSM profiles (BRPs) of protein families generated using a novel data mining approach.

BackgroundProtein families could be related to each other at broad levels that group them as superfamilies. These relationships are harder to detect at the sequence level due to high evolutionary divergence. Sequence searches are strongly directed and influenced by the best representatives of families that are viewed as starting points. PSSMs are useful approximations and mathematical representations of protein alignments, with wide array of applications in bioinformatics approaches like remote homology detection, protein family analysis, detection of new members and evolutionary modelling. Computational intensive searches have been performed using the neural network based sensitive sequence search method called FASSM to identify the Best Representative PSSMs for families reported in Pfam database version 22.ResultsWe designed a novel data mining approach for the assessment of individual sequences from a protein family to identify a single Best Representative PSSM profile (BRP) per protein family. Using the approach, a database of protein family-specific best representative PSSM profiles called 3PFDB has been developed. PSSM profiles in 3PFDB are curated using performance of individual sequence as a reference in a rigorous scoring and coverage analysis approach using FASSM. We have assessed the suitability of 10, 85,588 sequences derived from seed or full alignments reported in Pfam database (Version 22). Coverage analysis using FASSM method is used as the filtering step to identify the best representative sequence, starting from full length or domain sequences to generate the final profile for a given family. 3PFDB is a collection of best representative PSSM profiles of 8,524 protein families from Pfam database.ConclusionAvailability of an approach to identify BRPs and a curated database of best representative PSI-BLAST derived PSSMs for 91.4% of current Pfam family will be a useful resource for the community to perform detailed and specific analysis using family-specific, best-representative PSSM profiles. 3PFDB can be accessed using the URL: http://caps.ncbs.res.in/3pfdb