A graph-based clustering method applied to protein sequences

The number of amino acid sequences is increasing very rapidly in the protein databases like Swiss-Prot, Uniprot, PIR and others, but the structure of only some amino acid sequences are found in the Protein Data Bank. Thus, an important problem in genomics is automatically clustering homologous protein sequences when only sequence information is available. Here, we use graph theoretic techniques for clustering amino acid sequences. A similarity graph is defined and clusters in that graph correspond to connected subgraphs. Cluster analysis seeks grouping of amino acid sequences into subsets based on distance or similarity score between pairs of sequences. Our goal is to find disjoint subsets, called clusters, such that two criteria are satisfied: homogeneity: sequences in the same cluster are highly similar to each other; and separation: sequences in different clusters have low similarity to each other. We tested our method on several subsets of SCOP (Structural Classification of proteins) database, a gold standard for protein structure classification. The results show that for a given set of proteins the number of clusters we obtained is close to the superfamilies in that set; there are fewer singeltons; and the method correctly groups most remote homologs.     

An efficient algorithm for large-scale detection of protein families

Detection of protein families in large databases is one of the principal research objectives in structural and functional genomics. Protein family classification can significantly contribute to the delineation of functional diversity of homologous proteins, the prediction of function based on domain architecture or the presence of sequence motifs as well as comparative genomics, providing valuable evolutionary insights. We present a novel approach called TRIBE-MCL for rapid and accurate clustering of protein sequences into families. The method relies on the Markov cluster (MCL) algorithm for the assignment of proteins into families based on precomputed sequence similarity information. This novel approach does not suffer from the problems that normally hinder other protein sequence clustering algorithms, such as the presence of multi-domain proteins, promiscuous domains and fragmented proteins. The method has been rigorously tested and validated on a number of very large databases, including SwissProt, InterPro, SCOP and the draft human genome. Our results indicate that the method is ideally suited to the rapid and accurate detection of protein families on a large scale. The method has been used to detect and categorise protein families within the draft human genome and the resulting families have been used to annotate a large proportion of human proteins.     

Quantitative sequence-function relationships in proteins based on gene ontology

Background  

The relationship between divergence of amino-acid sequence and divergence of function among homologous proteins is complex. The assumption that homologs share function – the basis of transfer of annotations in databases – must therefore be regarded with caution. Here, we present a quantitative study of sequence and function divergence, based on the Gene Ontology classification of function. We determined the relationship between sequence divergence and function divergence in 6828 protein families from the PFAM database. Within families there is a broad range of sequence similarity from very closely related proteins – for instance, orthologs in different mammals – to very distantly-related proteins at the limit of reliable recognition of homology.     

Predicting functional family of novel enzymes irrespective of sequence similarity: a statistical learning approach

The function of a protein that has no sequence homolog of known function is difficult to assign on the basis of sequence similarity. The same problem may arise for homologous proteins of different functions if one is newly discovered and the other is the only known protein of similar sequence. It is desirable to explore methods that are not based on sequence similarity. One approach is to assign functional family of a protein to provide useful hint about its function. Several groups have employed a statistical learning method, support vector machines (SVMs), for predicting protein functional family directly from sequence irrespective of sequence similarity. These studies showed that SVM prediction accuracy is at a level useful for functional family assignment. But its capability for assignment of distantly related proteins and homologous proteins of different functions has not been critically and adequately assessed. Here SVM is tested for functional family assignment of two groups of enzymes. One consists of 50 enzymes that have no homolog of known function from PSI-BLAST search of protein databases. The other contains eight pairs of homologous enzymes of different families. SVM correctly assigns 72% of the enzymes in the first group and 62% of the enzyme pairs in the second group, suggesting that it is potentially useful for facilitating functional study of novel proteins. A web version of our software, SVMProt, is accessible at http://jing.cz3.nus.edu.sg/cgi-bin/svmprot.cgi.     

Modeling the percolation of annotation errors in a database of protein sequences

Public sequence databases contain information on the sequence, structure and function of proteins. Genome sequencing projects have led to a rapid increase in protein sequence information, but reliable, experimentally verified, information on protein function lags a long way behind. To address this deficit, functional annotation in protein databases is often inferred by sequence similarity to homologous, annotated proteins, with the attendant possibility of error. Now, the functional annotation in these homologous proteins may itself have been acquired through sequence similarity to yet other proteins, and it is generally not possible to determine how the functional annotation of any given protein has been acquired. Thus the possibility of chains of misannotation arises, a process we term 'error percolation'. With some simple assumptions, we develop a dynamical probabilistic model for these misannotation chains. By exploring the consequences of the model for annotation quality it is evident that this iterative approach leads to a systematic deterioration of database quality.     

Statistically rigorous automated protein annotation

MOTIVATION: Assignment of putative protein functional annotation by comparative analysis using pre-defined experimental annotations is performed routinely by molecular biologists. The number and statistical significance of these assignments remains a challenge in this era of high-throughput proteomics. A combined statistical method that enables robust, automated protein annotation by reliably expanding existing annotation sets is described. An existing clustering scheme, based on relevant experimental information (e.g. sequence identity, keywords or gene expression data) is required. The method assigns new proteins to these clusters with a measure of reliability. It can also provide human reviewers with a reliability score for both new and previously classified proteins. RESULTS: A dataset of 27 000 annotated Protein Data Bank (PDB) polypeptide chains (of 36 000 chains currently in the PDB) was generated from 23 000 chains classified a priori. AVAILABILITY: PDB annotations and sample software implementation are freely accessible on the Web at http://pmr.sdsc.edu/go     

Reduction of protein sequence complexity by residue grouping

It is well known that there are some similarities among various naturally occurring amino acids. Thus, the complexity in protein systems could be reduced by sorting these amino acids with similarities into groups and then protein sequences can be simplified by reduced alphabets. This paper discusses how to group similar amino acids and whether there is a minimal amino acid alphabet by which proteins can be folded. Various reduced alphabets are obtained by reserving the maximal information for the simplified protein sequence compared with the parent sequence using global sequence alignment. With these reduced alphabets and simplified similarity matrices, we achieve recognition of the protein fold based on the similarity score of the sequence alignment. The coverage in dataset SCOP40 for various levels of reduction on the amino acid types is obtained, which is the number of homologous pairs detected by program BLAST to the number marked by SCOP40. For the reduced alphabets containing 10 types of amino acids, the ability to detect distantly related folds remains almost at the same level as that by the alphabet of 20 types of amino acids, which implies that 10 types of amino acids may be the degree of freedom for characterizing the complexity in proteins.     

Prediction of yeast protein-protein interaction network: insights from the Gene Ontology and annotations

A map of protein–protein interactions provides valuable insight into the cellular function and machinery of a proteome. By measuring the similarity between two Gene Ontology (GO) terms with a relative specificity semantic relation, here, we proposed a new method of reconstructing a yeast protein–protein interaction map that is solely based on the GO annotations. The method was validated using high-quality interaction datasets for its effectiveness. Based on a Z-score analysis, a positive dataset and a negative dataset for protein–protein interactions were derived. Moreover, a gold standard positive (GSP) dataset with the highest level of confidence that covered 78% of the high-quality interaction dataset and a gold standard negative (GSN) dataset with the lowest level of confidence were derived. In addition, we assessed four high-throughput experimental interaction datasets using the positives and the negatives as well as GSPs and GSNs. Our predicted network reconstructed from GSPs consists of 40753 interactions among 2259 proteins, and forms 16 connected components. We mapped all of the MIPS complexes except for homodimers onto the predicted network. As a result, ~35% of complexes were identified interconnected. For seven complexes, we also identified some nonmember proteins that may be functionally related to the complexes concerned. This analysis is expected to provide a new approach for predicting the protein–protein interaction maps from other completely sequenced genomes with high-quality GO-based annotations.     

Integration of phenotypic metadata and protein similarity in Archaea using a spectral bipartitioning approach

In order to simplify and meaningfully categorize large sets of protein sequence data, it is commonplace to cluster proteins based on the similarity of those sequences. However, it quickly becomes clear that the sequence flexibility allowed a given protein varies significantly among different protein families. The degree to which sequences are conserved not only differs for each protein family, but also is affected by the phylogenetic divergence of the source organisms. Clustering techniques that use similarity thresholds for protein families do not always allow for these variations and thus cannot be confidently used for applications such as automated annotation and phylogenetic profiling. In this work, we applied a spectral bipartitioning technique to all proteins from 53 archaeal genomes. Comparisons between different taxonomic levels allowed us to study the effects of phylogenetic distances on cluster structure. Likewise, by associating functional annotations and phenotypic metadata with each protein, we could compare our protein similarity clusters with both protein function and associated phenotype. Our clusters can be analyzed graphically and interactively online.     

Prediction of functional class of novel bacterial proteins without the use of sequence similarity by a statistical learning method

A substantial percentage of the putative protein-encoding open reading frames (ORFs) in bacterial genomes have no homolog of known function, and their function cannot be confidently assigned on the basis of sequence similarity. Methods not based on sequence similarity are needed and being developed. One method, SVMProt (http://jing.cz3.nus.edu.sg/cgi-bin/svmprot.cgi), predicts protein functional family irrespective of sequence similarity (Nucleic Acids Res. 2003;31:3692-3697). While it has been tested on a large number of proteins, its capability for non-homologous proteins has so far been evaluated for a relatively small number of proteins, and additional tests are needed to more fully assess SVMProt. In this work, 90 novel bacterial proteins (non-homologous to known proteins) are used to evaluate the capability of SVMProt. These proteins are such that none of their homologs are in the Swiss-Prot database, their functions not clearly described in the literature, and they themselves and their homologs are not included in the training sets of SVMProt. They represent proteins whose function cannot be confidently predicted by sequence similarity methods at present. The predicted functional class of 76.7% of each of these proteins shows various levels of consistency with the literature-described function, compared to the overall accuracy of 87% for the SVMProt functional class assignment of 34,582 proteins that have at least one homolog of known function. Our study suggests that SVMProt is capable of assigning functional class for novel bacterial proteins at a level not too much lower than that of sequence alignment methods for homologous proteins.     

Protein-protein interaction network-based detection of functionally similar proteins within species

Although functionally similar proteins across species have been widely studied, functionally similar proteins within species showing low sequence similarity have not been examined in detail. Identification of these proteins is of significant importance for understanding biological functions, evolution of protein families, progression of co-evolution, and convergent evolution and others which cannot be obtained by detection of functionally similar proteins across species. Here, we explored a method of detecting functionally similar proteins within species based on graph theory. After denoting protein-protein interaction networks using graphs, we split the graphs into subgraphs using the 1-hop method. Proteins with functional similarities in a species were detected using a method of modified shortest path to compare these subgraphs and to find the eligible optimal results. Using seven protein-protein interaction networks and this method, some functionally similar proteins with low sequence similarity that cannot detected by sequence alignment were identified. By analyzing the results, we found that, sometimes, it is difficult to separate homologous from convergent evolution. Evaluation of the performance of our method by gene ontology term overlap showed that the precision of our method was excellent.     

Clusters in alpha/beta barrel proteins: implications for protein structure, function, and folding: a graph theoretical approach

The alpha/beta barrel fold is adopted by most enzymes performing a variety of catalytic reactions, but with very low sequence similarity. In order to understand the stabilizing interactions important in maintaining the alpha/beta barrel fold, we have identified residue clusters in a dataset of 36 alpha/beta barrel proteins that have less than 10% sequence identity within themselves. A graph theoretical algorithm is used to identify backbone clusters. This approach uses the global information of the nonbonded interaction in the alpha/beta barrel fold for the clustering procedure. The nonbonded interactions are represented mathematically in the form of an adjacency matrix. On diagonalizing the adjacency matrix, clusters and cluster centers are obtained from the highest eigenvalue and its corresponding vector components. Residue clusters are identified in the strand regions forming the beta barrel and are topologically conserved in all 36 proteins studied. The residues forming the cluster in each of the alpha/beta protein are also conserved among the sequences belonging to the same family. The cluster centers are found to occur in the middle of the strands or in the C-terminal of the strands. In most cases, the residues forming the clusters are part of the active site or are located close to the active site. The folding nucleus of the alpha/beta fold is predicted based on hydrophobicity index evaluation of residues and identification of cluster centers. The predicted nucleation sites are found to occur mostly in the middle of the strands. Proteins 2001;43:103-112.     

Comparison of sequence-based and structure-based phylogenetic trees of homologous proteins: Inferences on protein evolution

Several studies based on the known three-dimensional (3-D) structures of proteins show that two homologous proteins with insignificant sequence similarity could adopt a common fold and may perform same or similar biochemical functions. Hence, it is appropriate to use similarities in 3-D structure of proteins rather than the amino acid sequence similarities in modelling evolution of distantly related proteins. Here we present an assessment of using 3-D structures in modelling evolution of homologous proteins. Using a dataset of 108 protein domain families of known structures with at least 10 members per family we present a comparison of extent of structural and sequence dissimilarities among pairs of proteins which are inputs into the construction of phylogenetic trees. We find that correlation between the structure-based dissimilarity measures and the sequence-based dissimilarity measures is usually good if the sequence similarity among the homologues is about 30% or more. For protein families with low sequence similarity among the members, the correlation coefficient between the sequence-based and the structure-based dissimilarities are poor. In these cases the structure-based dendrogram clusters proteins with most similar biochemical functional properties better than the sequence-similarity based dendrogram. In multi-domain protein families and disulphide-rich protein families the correlation coefficient for the match of sequence-based and structure-based dissimilarity (SDM) measures can be poor though the sequence identity could be higher than 30%. Hence it is suggested that protein evolution is best modelled using 3-D structures if the sequence similarities (SSM) of the homologues are very low.     

Distinguishing enzyme structures from non-enzymes without alignments

The ability to predict protein function from structure is becoming increasingly important as the number of structures resolved is growing more rapidly than our capacity to study function. Current methods for predicting protein function are mostly reliant on identifying a similar protein of known function. For proteins that are highly dissimilar or are only similar to proteins also lacking functional annotations, these methods fail. Here, we show that protein function can be predicted as enzymatic or not without resorting to alignments. We describe 1178 high-resolution proteins in a structurally non-redundant subset of the Protein Data Bank using simple features such as secondary-structure content, amino acid propensities, surface properties and ligands. The subset is split into two functional groupings, enzymes and non-enzymes. We use the support vector machine-learning algorithm to develop models that are capable of assigning the protein class. Validation of the method shows that the function can be predicted to an accuracy of 77% using 52 features to describe each protein. An adaptive search of possible subsets of features produces a simplified model based on 36 features that predicts at an accuracy of 80%. We compare the method to sequence-based methods that also avoid calculating alignments and predict a recently released set of unrelated proteins. The most useful features for distinguishing enzymes from non-enzymes are secondary-structure content, amino acid frequencies, number of disulphide bonds and size of the largest cleft. This method is applicable to any structure as it does not require the identification of sequence or structural similarity to a protein of known function.     

DDOMAIN: Dividing structures into domains using a normalized domain-domain interaction profile

Dividing protein structures into domains is proven useful for more accurate structural and functional characterization of proteins. Here, we develop a method, called DDOMAIN, that divides structure into DOMAINs using a normalized contact-based domain-domain interaction profile. Results of DDOMAIN are compared to AUTHORS annotations (domain definitions are given by the authors who solved protein structures), as well as to popular SCOP and CATH annotations by human experts and automatic programs. DDOMAIN's automatic annotations are most consistent with the AUTHORS annotations (90% agreement in number of domains and 88% agreement in both number of domains and at least 85% overlap in domain assignment of residues) if its three adjustable parameters are trained by the AUTHORS annotations. By comparison, the agreement is 83% (81% with at least 85% overlap criterion) between SCOP-trained DDOMAIN and SCOP annotations and 77% (73%) between CATH-trained DDOMAIN and CATH annotations. The agreement between DDOMAIN and AUTHORS annotations goes beyond single-domain proteins (97%, 82%, and 56% for single-, two-, and three-domain proteins, respectively). For an "easy" data set of proteins whose CATH and SCOP annotations agree with each other in number of domains, the agreement is 90% (89%) between "easy-set"-trained DDOMAIN and CATH/SCOP annotations. The consistency between SCOP-trained DDOMAIN and SCOP annotations is superior to two other recently developed, SCOP-trained, automatic methods PDP (protein domain parser), and DomainParser 2. We also tested a simple consensus method made of PDP, DomainParser 2, and DDOMAIN and a different version of DDOMAIN based on a more sophisticated statistical energy function. The DDOMAIN server and its executable are available in the services section on http://sparks.informatics.iupui.edu.     

Enzyme function less conserved than anticipated

The level of sequence similarity that implies similarity in protein structure is well established. Recently, many groups proposed thresholds for similarity in sequence implying similarity in enzymatic function. All previous results suggest the strong conservation of enzymatic function above levels of 50% pairwise sequence identity. Here, I argue that all groups substantially overestimated the conservation of enzyme function because their data sets were either too biased, or too small. An unbiased analysis suggested that less than 30% of the pair fragments above 50% sequence identity have entirely identical EC numbers. Another surprising finding was that even BLAST E-values below 10(-50) did not suffice to automatically transfer enzyme function without errors. As expected, most misclassifications originated from similarities in relatively short regions and/or from transferring annotations for different domains. Both problems cannot be corrected easily by adjusting the thresholds for automatic transfer of genome annotations. A score relating sequence identity to alignment length (distance from HSSP-threshold) outperformed statistical BLAST scores for high sequence similarity. In particular, the distance score allowed error-free transfer of enzyme function for the 10% most similar enzyme pairs. The results illustrated how difficult it is to assess the conservation of protein function and to guarantee error-free genome annotations, in general: sets with millions of pair comparisons might not suffice to arrive at statistically significant conclusions. In practice, the revised detailed estimates for the sequence conservation of enzyme function may provide important benchmarks for everyday sequence analysis and for more cautious automatic genome annotations.     

Resolving protein structure‐function‐binding site relationships from a binding site similarity network perspective

Functional annotation is seldom straightforward with complexities arising due to functional divergence in protein families or functional convergence between non‐homologous protein families, leading to mis‐annotations. An enzyme may contain multiple domains and not all domains may be involved in a given function, adding to the complexity in function annotation. To address this, we use binding site information from bound cognate ligands and catalytic residues, since it can help in resolving fold‐function relationships at a finer level and with higher confidence. A comprehensive database of 2,020 fold‐function‐binding site relationships has been systematically generated. A network‐based approach is employed to capture the complexity in these relationships, from which different types of associations are deciphered, that identify versatile protein folds performing diverse functions, same function associated with multiple folds and one‐to‐one relationships. Binding site similarity networks integrated with fold, function, and ligand similarity information are generated to understand the depth of these relationships. Apart from the observed continuity in the functional site space, network properties of these revealed versatile families with topologically different or dissimilar binding sites and structural families that perform very similar functions. As a case study, subtle changes in the active site of a set of evolutionarily related superfamilies are studied using these networks. Tracing of such similarities in evolutionarily related proteins provide clues into the transition and evolution of protein functions. Insights from this study will be helpful in accurate and reliable functional annotations of uncharacterized proteins, poly‐pharmacology, and designing enzymes with new functional capabilities. Proteins 2017; 85:1319–1335. © 2017 Wiley Periodicals, Inc.     

Database of homology-derived protein structures and the structural meaning of sequence alignment

The database of known protein three-dimensional structures can be significantly increased by the use of sequence homology, based on the following observations. (1) The database of known sequences, currently at more than 12,000 proteins, is two orders of magnitude larger than the database of known structures. (2) The currently most powerful method of predicting protein structures is model building by homology. (3) Structural homology can be inferred from the level of sequence similarity. (4) The threshold of sequence similarity sufficient for structural homology depends strongly on the length of the alignment. Here, we first quantify the relation between sequence similarity, structure similarity, and alignment length by an exhaustive survey of alignments between proteins of known structure and report a homology threshold curve as a function of alignment length. We then produce a database of homology-derived secondary structure of proteins (HSSP) by aligning to each protein of known structure all sequences deemed homologous on the basis of the threshold curve. For each known protein structure, the derived database contains the aligned sequences, secondary structure, sequence variability, and sequence profile. Tertiary structures of the aligned sequences are implied, but not modeled explicitly. The database effectively increases the number of known protein structures by a factor of five to more than 1800. The results may be useful in assessing the structural significance of matches in sequence database searches, in deriving preferences and patterns for structure prediction, in elucidating the structural role of conserved residues, and in modeling three-dimensional detail by homology.