Mining for Class-Specific Motifs in Protein Sequence Classification

Background: In protein sequence classification, identification of the sequence motifs or n-grams that can precisely discriminate between classes is a more interesting scientific question than the classification itself. A number of classification methods aim at accurate classification but fail to explain which sequence features indeed contribute to the accuracy. We hypothesize that sequences in lower denominations (n-grams) can be used to explore the sequence landscape and to identify class-specific motifs that discriminate between classes during classification. Discriminative n-grams are short peptide sequences that are highly frequent in one class but are either minimally present or absent in other classes. In this study, we present a new substitution-based scoring function for identifying discriminative n-grams that are highly specific to a class. Results: We present a scoring function based on discriminative n-grams that can effectively discriminate between classes. The scoring function, initially, harvests the entire set of 4- to 8-grams from the protein sequences of different classes in the dataset. Similar n-grams of the same size are combined to form new n-grams, where the similarity is defined by positive amino acid substitution scores in the BLOSUM62 matrix. Substitution has resulted in a large increase in the number of discriminatory n-grams harvested. Due to the unbalanced nature of the dataset, the frequencies of the n-grams are normalized using a dampening factor, which gives more weightage to the n-grams that appear in fewer classes and vice-versa. After the n-grams are normalized, the scoring function identifies discriminative 4- to 8-grams for each class that are frequent enough to be above a selection threshold. By mapping these discriminative n-grams back to the protein sequences, we obtained contiguous n-grams that represent short class-specific motifs in protein sequences. Our method fared well compared to an existing motif finding method known as Wordspy. We have validated our enriched set of class-specific motifs against the functionally important motifs obtained from the NLSdb, Prosite and ELM databases. We demonstrate that this method is very generic; thus can be widely applied to detect class-specific motifs in many protein sequence classification tasks. Conclusion: The proposed scoring function and methodology is able to identify class-specific motifs using discriminative n-grams derived from the protein sequences. The implementation of amino acid substitution scores for similarity detection, and the dampening factor to normalize the unbalanced datasets have significant effect on the performance of the scoring function. Our multipronged validation tests demonstrate that this method can detect class-specific motifs from a wide variety of protein sequence classes with a potential application to detecting proteome-specific motifs of different organisms.

Present perspectives on the automated classification of the G-protein coupled receptors (GPCRs) at the protein sequence level

The G-protein coupled receptors--or GPCRs--comprise simultaneously one of the largest and one of the most multi-functional protein families known to modern-day molecular bioscience. From a drug discovery and pharmaceutical industry perspective, the GPCRs constitute one of the most commercially and economically important groups of proteins known. The GPCRs undertake numerous vital metabolic functions and interact with a hugely diverse range of small and large ligands. Many different methodologies have been developed to efficiently and accurately classify the GPCRs. These range from motif-based techniques to machine learning as well as a variety of alignment-free techniques based on the physiochemical properties of sequences. We review here the available methodologies for the classification of GPCRs. Part of this work focuses on how we have tried to build the intrinsically hierarchical nature of sequence relations, implicit within the family, into an adaptive approach to classification. Importantly, we also allude to some of the key innate problems in developing an effective approach to classifying the GPCRs: the lack of sequence similarity between the six classes that comprise the GPCR family and the low sequence similarity to other family members evinced by many newly revealed members of the family.

Protein sequence threading, the alignment problem, and a two-step strategy

Conventionally, protein structure prediction via threading relies on some nonoptimal method to align a protein sequence to each member of a library of known structures. We show how a score function (force field) can be modified so as to allow the direct application of a dynamic programming algorithm to the problem. This involves an approximation whose damage can be minimized by an optimization process during score function parameter determination. The method is compared to sequence to structure alignments using a more conventional pair-wise score function and the frozen approximation. The new method produces results comparable to the frozen approximation, but is faster and has fewer adjustable parameters. It is also free of memory of the template's original amino acid sequence, and does not suffer from a problem of nonconvergence, which can be shown to occur with the frozen approximation. Alignments generated by the simplified score function can then be ranked using a second score function with the approximations removed. (C) 1999 John Wiley & Sons, Inc.

HAMAP in 2015: updates to the protein family classification and annotation system.

HAMAP (High-quality Automated and Manual Annotation of Proteins-available at http://hamap.expasy.org/) is a system for the automatic classification and annotation of protein sequences. HAMAP provides annotation of the same quality and detail as UniProtKB/Swiss-Prot, using manually curated profiles for protein sequence family classification and expert curated rules for functional annotation of family members. HAMAP data and tools are made available through our website and as part of the UniRule pipeline of UniProt, providing annotation for millions of unreviewed sequences of UniProtKB/TrEMBL. Here we report on the growth of HAMAP and updates to the HAMAP system since our last report in the NAR Database Issue of 2013. We continue to augment HAMAP with new family profiles and annotation rules as new protein families are characterized and annotated in UniProtKB/Swiss-Prot; the latest version of HAMAP (as of 3 September 2014) contains 1983 family classification profiles and 1998 annotation rules (up from 1780 and 1720). We demonstrate how the complex logic of HAMAP rules allows for precise annotation of individual functional variants within large homologous protein families. We also describe improvements to our web-based tool HAMAP-Scan which simplify the classification and annotation of sequences, and the incorporation of an improved sequence-profile search algorithm.

Decreasing the number of false positives in sequence classification

Abstract Background A large number of probabilistic models used in sequence analysis assign non-zero probability values to most input sequences. To decide when a given probability is sufficient the most common way is bayesian binary classification, where the probability of the model characterizing the sequence family of interest is compared to that of an alternative probability model. We can use as alternative model a null model. This is the scoring technique used by sequence analysis tools such as HMMER, SAM and INFERNAL. The most prevalent null models are position-independent residue distributions that include: the uniform distribution, genomic distribution, family-specific distribution and the target sequence distribution. This paper presents a study to evaluate the impact of the choice of a null model in the final result of classifications. In particular, we are interested in minimizing the number of false predictions in a classification. This is a crucial issue to reduce costs of biological validation. Results For all the tests, the target null model presented the lowest number of false positives, when using random sequences as a test. The study was performed in DNA sequences using GC content as the measure of content bias, but the results should be valid also for protein sequences. To broaden the application of the results, the study was performed using randomly generated sequences. Previous studies were performed on aminoacid sequences, using only one probabilistic model (HMM) and on a specific benchmark, and lack more general conclusions about the performance of null models. Finally, a benchmark test with P. falciparum confirmed these results. Conclusions Of the evaluated models the best suited for classification are the uniform model and the target model. However, the use of the uniform model presents a GC bias that can cause more false positives for candidate sequences with extreme compositional bias, a characteristic not described in previous studies. In these cases the target model is more dependable for biological validation due to its higher specificity.

A clustering method for robust and reliable large scale functional and structural protein sequence annotation

Bioinformatics, in the last few decades, has played a fundamental role to give sense to the huge amount of data produced. Obtained the complete sequence of a genome, the major problem of knowing as much as possible of its coding regions, is crucial. Protein sequence annotation is challenging and, due to the size of the problem, only computational approaches can provide a feasible solution. As it has been recently pointed out by the Critical Assessment of Function Annotations (CAFA), most accurate methods are those based on the transfer-by-homology approach and the most incisive contribution is given by cross-genome comparisons. In the present thesis it is described a non-hierarchical sequence clustering method for protein automatic large-scale annotation, called “The Bologna Annotation Resource Plus” (BAR+). The method is based on an all-against-all alignment of more than 13 millions protein sequences characterized by a very stringent metric. BAR+ can safely transfer functional features (Gene Ontology and Pfam terms) inside clusters by means of a statistical validation, even in the case of multi-domain proteins. Within BAR+ clusters it is also possible to transfer the three dimensional structure (when a template is available). This is possible by the way of cluster-specific HMM profiles that can be used to calculate reliable template-to-target alignments even in the case of distantly related proteins (sequence identity < 30%). Other BAR+ based applications have been developed during my doctorate including the prediction of Magnesium binding sites in human proteins, the ABC transporters superfamily classification and the functional prediction (GO terms) of the CAFA targets. Remarkably, in the CAFA assessment, BAR+ placed among the ten most accurate methods. At present, as a web server for the functional and structural protein sequence annotation, BAR+ is freely available at http://bar.biocomp.unibo.it/bar2.0.

Evidence for an episodic model of protein sequence evolution.

The evolution of protein function appears to involve alternating periods of conservative evolution and of relatively rapid change. Evidence for such episodic evolution, consistent with some theoretical expectations, comes from the application of increasingly sophisticated models of evolution to large sequence datasets. We present here some of the recent methods to detect functional shifts, using amino acid or codon models. Both provide evidence for punctual shifts in patterns of amino acid conservation, including the fixation of key changes by positive selection. Although a link to gene duplication, a presumed source of functional changes, has been difficult to establish, this episodic model appears to apply to a wide variety of proteins and organisms.

Evolution at Two Levels in Fire Ants: The Relationship between Patterns of Gene Expression and Protein Sequence Evolution.

Variation in protein sequence and gene expression each contribute to phenotypic diversity, and may be subject to similar selective pressures. Eusocial insects are particularly useful for investigating the evolutionary link between protein sequence and condition-dependent patterns of gene expression because gene expression plays a central role in determining differences between eusocial insect sexes and castes. We investigated the relationship between protein coding sequence evolution and gene expression patterns in the fire ants Solenopsis invicta, S. richteri, and their hybrids to gain greater insight into how selection jointly operates on gene expression and coding sequence. We found that genes with high expression variability within castes and sexes were frequently differentially expressed between castes and sexes, as well as between species and hybrids. These results indicate that genes showing high variation in expression in one context also tend to show high variation in expression in other contexts. Our analyses further revealed that variation in both intra- and interspecific gene expression was positively associated with rate of protein sequence evolution in Solenopsis. This suggests that selective constraints on a gene operate both at the level of protein sequence and at the level of gene expression regulation. Overall, our study provides one of the strongest demonstrations that selective constraints mediate both protein sequence evolution and gene expression variability across different biological contexts and timescales.

Prediction of crude protein and classification of the growth stage of wheat plant samples from NIR spectra

Near-infrared spectroscopy (NIRS) was used to analyse the crude protein content of dried and milled samples of wheat and to discriminate samples according to their stage of growth. A calibration set of 72 samples from three growth stages of wheat (tillering, heading and harvest) and a validation set of 28 samples was collected for this purpose. Principal components analysis (PCA) of the calibration set discriminated groups of samples according to the growth stage of the wheat. Based on these differences, a classification procedure (SIMCA) showed a very accurate classification of the validation set samples : all of them were successfully classified in each group using this procedure when both the residual and the leverage were used in the classification criteria. Looking only at the residuals all the samples were also correctly classified except one of tillering stage that was assigned to both tillering and heading stages. Finally, the determination of the crude protein content of these samples was considered in two ways: building up a global model for all the growth stages, and building up local models for each stage, separately. The best prediction results for crude protein were obtained using a global model for samples in the two first growth stages (tillering and heading), and using a local model for the harvest stage samples.

Using genetic algorithm to design protein sequence

In this work, genetic algorithms concepts along with a rotamer library for proteins side chains are used to optimize the tertiary structure of the hydrophobic core of Cytochrome b(562) starting from the known PDB structure of its backbone which is kept fixed while the side chains of the hydrophobic core are allowed to adopt the conformations present in the rotamer library. The atoms of the side chains forming the core interact via van der Waals energy. Besides the prediction of the native core structure, it is also suggested a set of different amino acid sequences for this core. Comparison between these new cores and the native are made in terms of their volumes, van der Waals energies values and the numbers of contacts made by the side chains forming the cores. This paper proves that genetic algorithms area efficient to design new sequence for the protein core. (C) 2007 Elsevier B.V. All rights reserved.

The SBASE protein domain library, release 8.0: a collection of annotated protein sequence segments

SBASE 8.0 is the eighth release of the SBASE library of protein domain sequences that contains 294 898 annotated structural, functional, ligand-binding and topogenic segments of proteins, cross-referenced to most major sequence databases and sequence pattern collections. The entries are clustered into over 2005 statistically validated domain groups (SBASE-A) and 595 non-validated groups (SBASE-B), provided with several WWW-based search and browsing facilities for online use. A domain-search facility was developed, based on non-parametric pattern recognition methods, including artificial neural networks. SBASE 8.0 is freely available by anonymous ‘ftp’ file transfer from ftp.icgeb.trieste.it. Automated searching of SBASE can be carried out with the WWW servers http://www.icgeb.trieste.it/sbase/ and http://sbase.abc.hu/sbase/.

Highly specific protein sequence motifs for genome analysis

We present a method for discovering conserved sequence motifs from families of aligned protein sequences. The method has been implemented as a computer program called emotif (http://motif.stanford.edu/emotif). Given an aligned set of protein sequences, emotif generates a set of motifs with a wide range of specificities and sensitivities. emotif also can generate motifs that describe possible subfamilies of a protein superfamily. A disjunction of such motifs often can represent the entire superfamily with high specificity and sensitivity. We have used emotif to generate sets of motifs from all 7,000 protein alignments in the blocks and prints databases. The resulting database, called identify (http://motif.stanford.edu/identify), contains more than 50,000 motifs. For each alignment, the database contains several motifs having a probability of matching a false positive that range from 10−10 to 10−5. Highly specific motifs are well suited for searching entire proteomes, while generating very few false predictions. identify assigns biological functions to 25–30% of all proteins encoded by the Saccharomyces cerevisiae genome and by several bacterial genomes. In particular, identify assigned functions to 172 of proteins of unknown function in the yeast genome.

Multiple DNA and protein sequence alignment based on segment-to-segment comparison.

In this paper, a new way to think about, and to construct, pairwise as well as multiple alignments of DNA and protein sequences is proposed. Rather than forcing alignments to either align single residues or to introduce gaps by defining an alignment as a path running right from the source up to the sink in the associated dot-matrix diagram, we propose to consider alignments as consistent equivalence relations defined on the set of all positions occurring in all sequences under consideration. We also propose constructing alignments from whole segments exhibiting highly significant overall similarity rather than by aligning individual residues. Consequently, we present an alignment algorithm that (i) is based on segment-to-segment comparison instead of the commonly used residue-to-residue comparison and which (ii) avoids the well-known difficulties concerning the choice of appropriate gap penalties: gaps are not treated explicity, but remain as those parts of the sequences that do not belong to any of the aligned segments. Finally, we discuss the application of our algorithm to two test examples and compare it with commonly used alignment methods. As a first example, we aligned a set of 11 DNA sequences coding for functional helix-loop-helix proteins. Though the sequences show only low overall similarity, our program correctly aligned all of the 11 functional sites, which was a unique result among the methods tested. As a by-product, the reading frames of the sequences were identified. Next, we aligned a set of ribonuclease H proteins and compared our results with alignments produced by other programs as reported by McClure et al. [McClure, M. A., Vasi, T. K. & Fitch, W. M. (1994) Mol. Biol. Evol. 11, 571-592]. Our program was one of the best scoring programs. However, in contrast to other methods, our protein alignments are independent of user-defined parameters.

CyBase: a database of cyclic protein sequence and structure

CyBase is a curated database and information source for backbone-cyclized proteins. The database incorporates naturally occurring cyclic proteins as well as synthetic derivatives, grafted analogues and acyclic permutants. The database provides a centralized repository of information on all aspects of cyclic protein biology and addresses issues pertaining to the management and searching of topologically circular sequences. The database is freely available at http://research.imb.uq.edu.au/cybase.