Scoring Matrices Research Articles

Exploring the sequence determinants of amyloid structure using position-specific scoring matrices

Protein aggregation results in beta-sheet-like assemblies that adopt either a variety of amorphous morphologies or ordered amyloid-like structures. These differences in structure also reflect biological differences; amyloid and amorphous beta-sheet aggregates have different chaperone affinities, accumulate in different cellular locations and are degraded by different mechanisms. Further, amyloid function depends entirely on a high intrinsic degree of order. Here we experimentally explored the sequence space of amyloid hexapeptides and used the derived data to build Waltz, a web-based tool that uses a position-specific scoring matrix to determine amyloid-forming sequences. Waltz allows users to identify and better distinguish between amyloid sequences and amorphous beta-sheet aggregates and allowed us to identify amyloid-forming regions in functional amyloids.

Multivariate analysis of reflectance spectra from propolis: Geographical variation in Romanian samples

The present study described reflectance spectroscopy as a suitable analytical tool to discriminate the floral origin of 39 Romanian propolis samples. Relevant differences between the UV-vis reflectance spectra of the investigated propolis samples within the 220-850nm spectral range were found. The results obtained applying cluster analysis, principal component analysis and linear discriminant analysis to the digitized data of zero order, zero order normalized and first order derivative spectra support the reliability of this technique. In addition, the application of the linear discriminant analysis to the score matrices corresponding to the first principal components appeared to be an illuminating solution. Generally, the samples have been assigned to two large groups in a good agreement with their vegetal sampling location, samples originating from predominant forest area and samples originating from meadows. Within the first group, two subgroups were identified according to the dominant type of the forest, deciduous or resinous, while within the last group three subgroups were found according to the extend and variety of the meadow.

Significant speedup of database searches with HMMs by search space reduction with PSSM family models

Motivation: Profile hidden Markov models (pHMMs) are currently the most popular modeling concept for protein families. They provide sensitive family descriptors, and sequence database searching with pHMMs has become a standard task in today's genome annotation pipelines. On the downside, searching with pHMMs is computationally expensive.Results: We propose a new method for efficient protein family classification and for speeding up database searches with pHMMs as is necessary for large-scale analysis scenarios. We employ simpler models of protein families called position-specific scoring matrices family models (PSSM-FMs). For fast database search, we combine full-text indexing, efficient exact p-value computation of PSSM match scores and fast fragment chaining. The resulting method is well suited to prefilter the set of sequences to be searched for subsequent database searches with pHMMs. We achieved a classification performance only marginally inferior to hmmsearch, yet, results could be obtained in a fraction of runtime with a speedup of >64-fold. In experiments addressing the method's ability to prefilter the sequence space for subsequent database searches with pHMMs, our method reduces the number of sequences to be searched with hmmsearch to only 0.80% of all sequences. The filter is very fast and leads to a total speedup of factor 43 over the unfiltered search, while retaining >99.5% of the original results. In a lossless filter setup for hmmsearch on UniProtKB/Swiss-Prot, we observed a speedup of factor 92.Availability: The presented algorithms are implemented in the program PoSSuMsearch2, available for download at http://bibiserv.techfak.uni-bielefeld.de/possumsearch2/.Contact: beckstette@zbh.uni-hamburg.deSupplementary information: Supplementary data are available at Bioinformatics online.

CCHMM_PROF: a HMM-based coiled-coil predictor with evolutionary information

The widespread coiled-coil structural motif in proteins is known to mediate a variety of biological interactions. Recognizing a coiled-coil containing sequence and locating its coiled-coil domains are key steps towards the determination of the protein structure and function. Different tools are available for predicting coiled-coil domains in protein sequences, including those based on position-specific score matrices and machine learning methods. In this article, we introduce a hidden Markov model (CCHMM_PROF) that exploits the information contained in multiple sequence alignments (profiles) to predict coiled-coil regions. The new method discriminates coiled-coil sequences with an accuracy of 97% and achieves a true positive rate of 79% with only 1% of false positives. Furthermore, when predicting the location of coiled-coil segments in protein sequences, the method reaches an accuracy of 80% at the residue level and a best per-segment and per-protein efficiency of 81% and 80%, respectively. The results indicate that CCHMM_PROF outperforms all the existing tools and can be adopted for large-scale genome annotation. The dataset is available at http://www.biocomp.unibo.it/ approximately lisa/coiled-coils. The predictor is freely available at http://gpcr.biocomp.unibo.it/cgi/predictors/cchmmprof/pred_cchmmprof.cgi. piero@biocomp.unibo.it.

A new taxonomy-based protein fold recognition approach based on autocross-covariance transformation

Fold recognition is an important step in protein structure and function prediction. Traditional sequence comparison methods fail to identify reliable homologies with low sequence identity, while the taxonomic methods are effective alternatives, but their prediction accuracies are around 70%, which are still relatively low for practical usage. In this study, a simple and powerful method is presented for taxonomic fold recognition, which combines support vector machine (SVM) with autocross-covariance (ACC) transformation. The evolutionary information represented in the form of position-specific score matrices is converted into a series of fixed-length vectors by ACC transformation and these vectors are then input to a SVM classifier for fold recognition. The sequence-order effect can be effectively captured by this scheme. Experiments are performed on the widely used D-B dataset and the corresponding extended dataset, respectively. The proposed method, called ACCFold, gets an overall accuracy of 70.1% on the D-B dataset, which is higher than major existing taxonomic methods by 2-14%. Furthermore, the method achieves an overall accuracy of 87.6% on the extended dataset, which surpasses major existing taxonomic methods by 9-17%. Additionally, our method obtains an overall accuracy of 80.9% for 86-folds and 77.2% for 199-folds. These results demonstrate that the ACCFold method provides the state-of-the-art performance for taxonomic fold recognition. The source code for ACC transformation is freely available at http://www.iipl.fudan.edu.cn/demo/accpkg.html.

Joint Procrustes Analysis for Simultaneous Nonsingular Transformation of Component Score and Loading Matrices

In component analysis solutions, post-multiplying a component score matrix by a nonsingular matrix can be compensated by applying its inverse to the corresponding loading matrix. To eliminate this indeterminacy on nonsingular transformation, we propose Joint Procrustes Analysis (JPA) in which component score and loading matrices are simultaneously transformed so that the former matrix matches a target score matrix and the latter matches a target loading matrix. The loss function of JPA is a function of the nonsingular transformation matrix and its inverse, and is difficult to minimize straightforwardly. To deal with this difficulty, we reparameterize those matrices by their singular value decomposition, which reduces the minimization to alternately solving quartic equations and performing the existing multivariate procedures. This algorithm is assessed in a simulation study. We further extend JPA for cases where targets are linear functions of unknown parameters. We also discuss how the application of JPA can be extended in different fields.

ALIGN_MTX—An optimal pairwise textual sequence alignment program, adapted for using in sequence-structure alignment

The presented program ALIGN_MTX makes alignment of two textual sequences with an opportunity to use any several characters for the designation of sequence elements and arbitrary user substitution matrices. It can be used not only for the alignment of amino acid and nucleotide sequences but also for sequence-structure alignment used in threading, amino acid sequence alignment, using preliminary known PSSM matrix, and in other cases when alignment of biological or non-biological textual sequences is required. This distinguishes it from the majority of similar alignment programs that make, as a rule, alignment only of amino acid or nucleotide sequences represented as a sequence of single alphabetic characters. ALIGN_MTX is presented as downloadable zip archive at http://www.imbbp.org/software/ALIGN_MTX/ and available for free use. As application of using the program, the results of comparison of different types of substitution matrix for alignment quality in distantly related protein pair sets were presented. Threading matrix SORDIS, based on side-chain orientation in relation to hydrophobic core centers with evolutionary change-based substitution matrix BLOSUM and using multiple sequence alignment information position-specific score matrices (PSSM) were taken for test alignment accuracy. The best performance shows PSSM matrix, but in the reduced set with lower sequence similarity threading matrix SORDIS shows the same performance and it was shown that combined potential with SORDIS and PSSM can improve alignment quality in evolutionary distantly related protein pairs.

Presence of HIV-1 R5 Viruses in Cerebrospinal Fluid Even in Patients Harboring R5X4/X4 Viruses in Plasma

HIV-1 viruses have the ability to use CCR5 or CXCR4 coreceptors either solely (R5 or X4) or in combination (R5X4). The CCR5 antagonists block HIV entry into the cell and are specifically active against HIV-1 R5 strains. The objectives of this study were to investigate the predicted tropism of viruses present in paired cerebrospinal fluid (CSF) and plasma in a group of 22 HIV-1-infected patients with neurological disorders and to search for eventual discordance of predicted virus tropism between both compartments. Paired CSF and plasma samples were selected from subjects harboring neurological disorders. V3 env was amplified and bulk sequenced, and HIV-1 coreceptor usage was determined from the V3 env region sequence by Geno2Pheno and position-specific scoring matrices (PSSM) algorithms. The majority of subjects (19 of 22) had concordant virus predicted tropism in both compartments. All patients having R5-tropic viruses in plasma had R5-tropic viruses in CSF (17 of 22). Patients having R5X4/X4-tropic viruses in plasma could have R5X4/X4-using (2 of 22) or R5-tropic viruses (3 of 22) in CSF. The case of R5-tropic viruses in plasma and R5X4/X4-tropic viruses in CSF was never observed. The majority of these patients have R5-using viruses in CSF that is mainly concordant with the predicted tropism in plasma. However, R5X4/X4 tropism in plasma does not necessarily mean the same predicted tropism in CSF compartment. Then, clinical therapeutic trials testing the clinical response to the CCR5 antagonists in patients with neurological disorders could be envisaged to analyze the effects of this therapeutic class in this setting.

The p53HMM algorithm: using profile hidden markov models to detect p53-responsive genes

BackgroundA computational method (called p53HMM) is presented that utilizes Profile Hidden Markov Models (PHMMs) to estimate the relative binding affinities of putative p53 response elements (REs), both p53 single-sites and cluster-sites. These models incorporate a novel "Corresponded Baum-Welch" training algorithm that provides increased predictive power by exploiting the redundancy of information found in the repeated, palindromic p53-binding motif. The predictive accuracy of these new models are compared against other predictive models, including position specific score matrices (PSSMs, or weight matrices). We also present a new dynamic acceptance threshold, dependent upon a putative binding site's distance from the Transcription Start Site (TSS) and its estimated binding affinity. This new criteria for classifying putative p53-binding sites increases predictive accuracy by reducing the false positive rate.ResultsTraining a Profile Hidden Markov Model with corresponding positions matching a combined-palindromic p53-binding motif creates the best p53-RE predictive model. The p53HMM algorithm is available on-line: ConclusionUsing Profile Hidden Markov Models with training methods that exploit the redundant information of the homotetramer p53 binding site provides better predictive models than weight matrices (PSSMs). These methods may also boost performance when applied to other transcription factor binding sites.

SpliceAid: a database of experimental RNA target motifs bound by splicing proteins in humans

The correct post-transcriptional RNA processing is finely regulated by RNA-binding proteins. Unfortunately, there is little experimental information on target RNA sequences of RNA-binding proteins and moreover such experimentally derived target sequences are annotated in a compact form by the score matrices that overestimate the number of possible recognized sequences. We carried out an exhaustive hand curated literature search to create a database, SpliceAid, collecting all the experimentally assessed target RNA sequences that are bound by splicing proteins in humans. We built a web resource, database driven, to easy query SpliceAid and give back the results by an accurate and dynamic graphic representation. SpliceAid database is freely accessible at http://www.introni.it/splicing.html.

Two new extensions of principal component transform to compute a PLS2 model between two wide matrices: PCT-PLS2 and segmented PCT-PLS2

The progress of analytical techniques has led to the possibility of acquiring a large number of data for each analysed sample. Moreover, the application of pre-treatment methods such as Contrast greatly increases the number of variables, yielding (very) wide data matrices. The computation of a PLS2 model between two such matrices may be slowed down, or made impossible, because of computer memory problems. The method presented in this article proposes an algorithm to solve this problem, and to enable the computation of a PLS2 model between two matrices containing a large number of variables. To do this, the PLS2 model is computed between the score matrices obtained by a PCA on each original matrix separately. After PLS2, a back-transformation to the original space is possible, and leads to results identical to those which would have been obtained in the original space. The method can be later extended, by segmenting the matrices, and computing the PC transform on each segment, before concatenating all the resulting score matrices and computing the PLS2 model on the obtained matrices.

Prediction of DNA-binding residues from protein sequence information using random forests

BackgroundProtein-DNA interactions are involved in many biological processes essential for cellular function. To understand the molecular mechanism of protein-DNA recognition, it is necessary to identify the DNA-binding residues in DNA-binding proteins. However, structural data are available for only a few hundreds of protein-DNA complexes. With the rapid accumulation of sequence data, it becomes an important but challenging task to accurately predict DNA-binding residues directly from amino acid sequence data.ResultsA new machine learning approach has been developed in this study for predicting DNA-binding residues from amino acid sequence data. The approach used both the labelled data instances collected from the available structures of protein-DNA complexes and the abundant unlabeled data found in protein sequence databases. The evolutionary information contained in the unlabeled sequence data was represented as position-specific scoring matrices (PSSMs) and several new descriptors. The sequence-derived features were then used to train random forests (RFs), which could handle a large number of input variables and avoid model overfitting. The use of evolutionary information was found to significantly improve classifier performance. The RF classifier was further evaluated using a separate test dataset, and the predicted DNA-binding residues were examined in the context of three-dimensional structures.ConclusionThe results suggest that the RF-based approach gives rise to more accurate prediction of DNA-binding residues than previous studies. A new web server called BindN-RF has thus been developed to make the RF classifier accessible to the biological research community.

PSI-BLAST pseudocounts and the minimum description length principle.

Position specific score matrices (PSSMs) are derived from multiple sequence alignments to aid in the recognition of distant protein sequence relationships. The PSI-BLAST protein database search program derives the column scores of its PSSMs with the aid of pseudocounts, added to the observed amino acid counts in a multiple alignment column. In the absence of theory, the number of pseudocounts used has been a completely empirical parameter. This article argues that the minimum description length principle can motivate the choice of this parameter. Specifically, for realistic alignments, the principle supports the practice of using a number of pseudocounts essentially independent of alignment size. However, it also implies that more highly conserved columns should use fewer pseudocounts, increasing the inter-column contrast of the implied PSSMs. A new method for calculating pseudocounts that significantly improves PSI-BLAST's; retrieval accuracy is now employed by default.

The value of position-specific scoring matrices for assessment of protein allegenicity

BackgroundBioinformatics tools are commonly used for assessing potential protein allergenicity. While these methods have achieved good accuracies for highly conserved sequences, they are less effective when the overall similarity is low. In this study, we assessed the feasibility of using position-specific scoring matrices as a basis for predicting potential allergenicity in proteins.ResultsTwo simple methods for predicting potential allergenicity in proteins, based on general and group-specific allergen profiles, are presented. Testing results indicate that the performances of both methods are comparable to the best results of other methods. The group-specific profile approach, with a sensitivity of 84.04% and specificity of 96.52%, gives similar results as those obtained using the general profile approach (sensitivity = 82.45%, specificity = 96.92%).ConclusionWe show that position-specific scoring matrices are highly promising for constructing computational models suitable for allergenicity assessment. These data suggest it may be possible to apply a targeted approach for allergenicity assessment based on the profiles of allergens of interest.

Optimal neighborhood indexing for protein similarity search

BackgroundSimilarity inference, one of the main bioinformatics tasks, has to face an exponential growth of the biological data. A classical approach used to cope with this data flow involves heuristics with large seed indexes. In order to speed up this technique, the index can be enhanced by storing additional information to limit the number of random memory accesses. However, this improvement leads to a larger index that may become a bottleneck. In the case of protein similarity search, we propose to decrease the index size by reducing the amino acid alphabet.ResultsThe paper presents two main contributions. First, we show that an optimal neighborhood indexing combining an alphabet reduction and a longer neighborhood leads to a reduction of 35% of memory involved into the process, without sacrificing the quality of results nor the computational time. Second, our approach led us to develop a new kind of substitution score matrices and their associated e-value parameters. In contrast to usual matrices, these matrices are rectangular since they compare amino acid groups from different alphabets. We describe the method used for computing those matrices and we provide some typical examples that can be used in such comparisons. Supplementary data can be found on the website .ConclusionWe propose a practical index size reduction of the neighborhood data, that does not negatively affect the performance of large-scale search in protein sequences. Such an index can be used in any study involving large protein data. Moreover, rectangular substitution score matrices and their associated statistical parameters can have applications in any study involving an alphabet reduction.

RNAalifold: improved consensus structure prediction for RNA alignments

BackgroundThe prediction of a consensus structure for a set of related RNAs is an important first step for subsequent analyses. RNAalifold, which computes the minimum energy structure that is simultaneously formed by a set of aligned sequences, is one of the oldest and most widely used tools for this task. In recent years, several alternative approaches have been advocated, pointing to several shortcomings of the original RNAalifold approach.ResultsWe show that the accuracy of RNAalifold predictions can be improved substantially by introducing a different, more rational handling of alignment gaps, and by replacing the rather simplistic model of covariance scoring with more sophisticated RIBOSUM-like scoring matrices. These improvements are achieved without compromising the computational efficiency of the algorithm. We show here that the new version of RNAalifold not only outperforms the old one, but also several other tools recently developed, on different datasets.ConclusionThe new version of RNAalifold not only can replace the old one for almost any application but it is also competitive with other approaches including those based on SCFGs, maximum expected accuracy, or hierarchical nearest neighbor classifiers.

CC+: a relational database of coiled-coil structures

We introduce the CC+ Database, a detailed, searchable repository of coiled-coil assignments, which is freely available at http://coiledcoils.chm.bris.ac.uk/ccplus. Coiled coils were identified using the program SOCKET, which locates coiled coils based on knobs-into-holes packing of side chains between α-helices. A method for determining the overall sequence identity of coiled-coil sequences was introduced to reduce statistical bias inherent in coiled-coil data sets. There are two points of entry into the CC+ Database: the ‘Periodic Table of Coiled-coil Structures’, which presents a graphical path through coiled-coil space based on manually validated data, and the ‘Dynamic Interface’, which allows queries of the database at different levels of complexity and detail. The latter entry level, which is the focus of this article, enables the efficient and rapid compilation of subsets of coiled-coil structures. These can be created and interrogated with increasingly sophisticated pull-down, keyword and sequence-based searches to return detailed structural and sequence information. Also provided are means for outputting the retrieved coiled-coil data in various formats, including PyMOL and RasMol scripts, and Position-Specific Scoring Matrices (or amino-acid profiles), which may be used, for example, in protein-structure prediction.

Using RSAT to scan genome sequences for transcription factor binding sites and cis-regulatory modules

This protocol shows how to detect putative cis-regulatory elements and regions enriched in such elements with the regulatory sequence analysis tools (RSAT) web server (http://rsat.ulb.ac.be/rsat/). The approach applies to known transcription factors, whose binding specificity is represented by position-specific scoring matrices, using the program matrix-scan. The detection of individual binding sites is known to return many false predictions. However, results can be strongly improved by estimating P value, and by searching for combinations of sites (homotypic and heterotypic models). We illustrate the detection of sites and enriched regions with a study case, the upstream sequence of the Drosophila melanogaster gene even-skipped. This protocol is also tested on random control sequences to evaluate the reliability of the predictions. Each task requires a few minutes of computation time on the server. The complete protocol can be executed in about one hour.

The effectiveness of position- and composition-specific gap costs for protein similarity searches

Motivation: The flexibility in gap cost enjoyed by hidden Markov models (HMMs) is expected to afford them better retrieval accuracy than position-specific scoring matrices (PSSMs). We attempt to quantify the effect of more general gap parameters by separately examining the influence of position- and composition-specific gap scores, as well as by comparing the retrieval accuracy of the PSSMs constructed using an iterative procedure to that of the HMMs provided by Pfam and SUPERFAMILY, curated ensembles of multiple alignments.Results: We found that position-specific gap penalties have an advantage over uniform gap costs. We did not explore optimizing distinct uniform gap costs for each query. For Pfam, PSSMs iteratively constructed from seeds based on HMM consensus sequences perform equivalently to HMMs that were adjusted to have constant gap transition probabilities, albeit with much greater variance. We observed no effect of composition-specific gap costs on retrieval performance. These results suggest possible improvements to the PSI-BLAST protein database search program.Availability: The scripts for performing evaluations are available upon request from the authors.Contact:yyu@ncbi.nlm.nih.gov

A versatile strategy to define the phosphorylation preferences of plant protein kinases and screen for putative substrates

Most signaling networks are regulated by reversible protein phosphorylation. The specificity of this regulation depends in part on the capacity of protein kinases to recognize and efficiently phosphorylate particular sequence motifs in their substrates. Sequenced plant genomes potentially encode over than 1000 protein kinases, representing 4% of the proteins, twice the proportion found in humans. This plethora of plant kinases requires the development of high-throughput strategies to identify their substrates. In this study, we have implemented a semi-degenerate peptide array screen to define the phosphorylation preferences of four kinases from Arabidopsis thaliana that are representative of the plant calcium-dependent protein kinase and Snf1-related kinase superfamily. We converted these quantitative data into position-specific scoring matrices to identify putative substrates of these kinases in silico in protein sequence databases. Our data show that these kinases display related but nevertheless distinct phosphorylation motif preferences, suggesting that they might share common targets but are likely to have specific substrates. Our analysis also reveals that a conserved motif found in the stress-related dehydrin protein family may be targeted by the SnRK2-10 kinase. Our results indicate that semi-degenerate peptide array screening is a versatile strategy that can be used on numerous plant kinases to facilitate identification of their substrates, and therefore represents a valuable tool to decipher phosphorylation-regulated signaling networks in plants.