Search

Supervised machine learning is widely used in biology and deserves more scrutiny. We present a set of community-wide recommendations (DOME) aiming to help establish standards of supervised machine learning validation in biology. Formulated as questions, the DOME recommendations improve the assessment and reproducibility of papers when included as supplementary material. The work of the Machine Learning Focus Group was funded by ELIXIR, the research infrastructure for life-science data. IW was funded by the A*STAR Career Development Award (project no. C210112057) from the Agency for Science, Technology and Research (A*STAR), Singapore. D.F. was supported by Estonian Research Council grants (PRG1095, PSG59 and ERA-NET TRANSCAN-2 (BioEndoCar)); Project No 2014-2020.4.01.16-0271, ELIXIR and the European Regional Development Fund through EXCITE Center of Excellence. S.C.E.T. has received funding from the European Union’s Horizon 2020 research and innovation programme under Marie Skłodowska-Curie Grant agreements No. 778247 and No. 823886, and Italian Ministry of University and Research PRIN 2017 grant 2017483NH8. Peer Reviewed "Article signat per 8 autors més 28 autors/es de l' ELIXIR Machine Learning Focus Group: Emidio Capriotti, Rita Casadio, Salvador Capella-Gutierrez, Davide Cirillo, Alessio Del Conte, Alexandros C. Dimopoulos, Victoria Dominguez Del Angel, Joaquin Dopazo, Piero Fariselli, José Maria Fernández, Florian Huber, Anna Kreshuk, Tom Lenaerts, Pier Luigi Martelli, Arcadi Navarro, Pilib Ó Broin, Janet Piñero, Damiano Piovesan, Martin Reczko, Francesco Ronzano, Venkata Satagopam, Castrense Savojardo, Vojtech Spiwok, Marco Antonio Tangaro, Giacomo Tartari, David Salgado, Alfonso Valencia & Federico Zambelli"

Frataxin (FXN) is a highly conserved protein found in prokaryotes and eukaryotes that is required for efficient regulation of cellular iron homeostasis. Experimental evidence associates amino acid substitutions of the FXN to Friedreich Ataxia, a neurodegenerative disorder. Recently, new thermodynamic experiments have been performed to study the impact of somatic variations identified in cancer tissues on protein stability. The Critical Assessment of Genome Interpretation (CAGI) data provider at the University of Rome measured the unfolding free energy of a set of variants (FXN challenge data set) with far-UV circular dichroism and intrinsic fluorescence spectra. These values have been used to calculate the change in unfolding free energy between the variant and wild-type proteins at zero concentration of denaturant ( Δ Δ G H 2 O ) . The FXN challenge data set, composed of eight amino acid substitutions, was used to evaluate the performance of the current computational methods for predicting the Δ Δ G H 2 O value associated with the variants and to classify them as destabilizing and not destabilizing. For the fifth edition of CAGI, six independent research groups from Asia, Australia, Europe, and North America submitted 12 sets of predictions from different approaches. In this paper, we report the results of our assessment and discuss the limitations of the tested algorithms.

Protein sequences, directly translated from genomic data, need functional and structural annotation. Together with molecular function and biological process, subcellular localization is an important feature necessary for understanding the protein role and the compartment where the mature protein is active. In the case of mitochondrial proteins, their precursor sequences translated by the ribosome machinery include specific patterns from which it is possible not only to recognize their final destination within the organelle but also which of the mitochondrial subcompartments the protein is intended for. Four compartments are routinely discriminated, including the inner and the outer membranes, the intermembrane space, and the matrix. Here we discuss to which extent it is feasible to develop computational methods for detecting mitochondrial targeting peptides in the precursor sequence and todiscriminate their final destination in the organelle. We benchmark two of our methods on the general task of recognizing human mitochondrial proteins endowed with an experimentally characterized targeting peptide (TPpred3) and predicting which submitochondrial compartment is the final destination (DeepMito). We describe how to adopt our web servers in order to discriminate which human proteins are endowed with mitochondrial targeting peptides, the position of cleavage sites, and which submitochondrial compartment are intended for. By this, we add some other 1788 human proteins to the 450 ones already manually annotated in UniProt with a mitochondrial targeting peptide, providing for each of them also the characterization of the suborganellar localization.

Graphical abstract, Protein stability predictions are becoming essential in medicine to develop novel immunotherapeutic agents and for drug discovery. Despite the large number of computational approaches for predicting the protein stability upon mutation, there are still critical unsolved problems: 1) the limited number of thermodynamic measurements for proteins provided by current databases; 2) the large intrinsic variability of ΔΔG values due to different experimental conditions; 3) biases in the development of predictive methods caused by ignoring the anti-symmetry of ΔΔG values between mutant and native protein forms; 4) over-optimistic prediction performance, due to sequence similarity between proteins used in training and test datasets. Here, we review these issues, highlighting new challenges required to improve current tools and to achieve more reliable predictions. In addition, we provide a perspective of how these methods will be beneficial for designing novel precision medicine approaches for several genetic disorders caused by mutations, such as cancer and neurodegenerative diseases.

Motivation: Targeting peptides are the most important signal controlling the import of nuclear encoded proteins into mitochondria and plastids. In the lack of experimental information, their prediction is an essential step when proteomes are annotated for inferring both the localization and the sequence of mature proteins. Results: We developed TPpred a new predictor of organelle-targeting peptides based on Grammatical-Restrained Hidden Conditional Random Fields. TPpred is trained on a non-redundant dataset of proteins where the presence of a target peptide was experimentally validated, comprising 297 sequences. When tested on the 297 positive and some other 8010 negative examples, TPpred outperformed available methods in both accuracy and Matthews correlation index (96% and 0.58, respectively). Given its very low–false-positive rate (3.0%), TPpred is, therefore, well suited for large-scale analyses at the proteome level. We predicted that from ∼4 to 9% of the sequences of human, Arabidopsis thaliana and yeast proteomes contain targeting peptides and are, therefore, likely to be localized in mitochondria and plastids. TPpred predictions correlate to a good extent with the experimental annotation of the subcellular localization, when available. TPpred was also trained and tested to predict the cleavage site of the organelle-targeting peptide: on this task, the average error of TPpred on mitochondrial and plastidic proteins is 7 and 15 residues, respectively. This value is lower than the error reported by other methods currently available. Availability: The TPpred datasets are available at http://biocomp.unibo.it/∼valentina/TPpred/. TPpred is available on request from the authors. Contact: gigi@biocomp.unibo.it Supplementary information: Supplementary data are available at Bioinformatics online.

We introduce BAR-PLUS (BAR(+)), a web server for functional and structural annotation of protein sequences. BAR(+) is based on a large-scale genome cross comparison and a non-hierarchical clustering procedure characterized by a metric that ensures a reliable transfer of features within clusters. In this version, the method takes advantage of a large-scale pairwise sequence comparison of 13,495,736 protein chains also including 988 complete proteomes. Available sequence annotation is derived from UniProtKB, GO, Pfam and PDB. When PDB templates are present within a cluster (with or without their SCOP classification), profile Hidden Markov Models (HMMs) are computed on the basis of sequence to structure alignment and are cluster-associated (Cluster-HMM). Therefrom, a library of 10,858 HMMs is made available for aligning even distantly related sequences for structural modelling. The server also provides pairwise query sequence-structural target alignments computed from the correspondent Cluster-HMM. BAR(+) in its present version allows three main categories of annotation: PDB [with or without SCOP (*)] and GO and/or Pfam; PDB (*) without GO and/or Pfam; GO and/or Pfam without PDB (*) and no annotation. Each category can further comprise clusters where GO and Pfam functional annotations are or are not statistically significant. BAR(+) is available at http://bar.biocomp.unibo.it/bar2.0.

In this paper we aim at investigating possible correlations between the number of putative interaction patches of a given protein, as inferred by an algorithm that we have developed, and its degree (number of edges of the protein node in a protein interaction network). We focus on the human cell cycle that, as compared with other biological processes, comprises the largest number of proteins whose structure is known at atomic resolution both as monomers and as interacting complexes. For predicting interaction patches we specifically develop a HM-SVM based method reaching 71% overall accuracy with a correlation coefficient value equal to 0.43 on a non redundant set of protein complexes. To test the biological meaning of our predictions, we also explore whether interacting patches contain energetically important residues and/or disease related mutations and find that predicted patches are endowed with both features. Based on this, we propose that mapping the protein with all the predicted interaction patches bridges the molecule to the interactome at the cell level. To test our hypothesis we downloaded interaction data from interaction data bases and find that the number of predicted interaction patches significantly correlates (Pearson correlation value > 0.3) with the number of the known interactions (edges) per protein in the human interactome, as contained in MINT and IntAct. We also show that the correlation increases (Pearson correlation value > 0.5) when the subcellular co-localization and the co-expression levels of the interacting partners are taken into account.

Motivation: Searching for structural similarity is a key issue of protein functional annotation. The maximum contact map overlap (CMO) is one of the possible measures of protein structure similarity. Exact and approximate methods known to optimize the CMO are computationally expensive and this hampers their applicability to large-scale comparison of protein structures. Results: In this article, we describe a heuristic algorithm (Al-Eigen) for finding a solution to the CMO problem. Our approach relies on the approximation of contact maps by eigendecomposition. We obtain good overlaps of two contact maps by computing the optimal global alignment of few principal eigenvectors. Our algorithm is simple, fast and its running time is independent of the amount of contacts in the map. Experimental testing indicates that the algorithm is comparable to exact CMO methods in terms of the overlap quality, to structural alignment methods in terms of structure similarity detection and it is fast enough to be suited for large-scale comparison of protein structures. Furthermore, our preliminary tests indicates that it is quite robust to noise, which makes it suitable for structural similarity detection also for noisy and incomplete contact maps. Availability: Available at http://bioinformatics.cs.unibo.it/Al-Eigen Contact: dilena@cs.unibo.it Supplementary information: Supplementary data are available at Bioinformatics online.

Extensive amounts of information about protein sequences are becoming available, as demonstrated by the over 79 million entries in the UniProt database. Yet, it is still challenging to obtain proteome-wide experimental information on the structural properties associated with these sequences. Fast computational predictors of secondary structure and of intrinsic disorder of proteins have been developed in order to bridge this gap. These two types of predictions, however, have remained largely separated, often preventing a clear characterization of the structure and dynamics of proteins. Here, we introduce a computational method to predict secondary-structure populations from amino acid sequences, which simultaneously characterizes structure and disorder in a unified statistical mechanics framework. To develop this method, called s2D, we exploited recent advances made in the analysis of NMR chemical shifts that provide quantitative information about the probability distributions of secondary-structure elements in disordered states. The results that we discuss show that the s2D method predicts secondary-structure populations with an average error of about 14%. A validation on three datasets of mostly disordered, mostly structured and partly structured proteins, respectively, shows that its performance is comparable to or better than that of existing predictors of intrinsic disorder and of secondary structure. These results indicate that it is possible to perform rapid and quantitative sequence-based characterizations of the structure and dynamics of proteins through the predictions of the statistical distributions of their ordered and disordered regions.

Background In the genomic era a key issue is protein annotation, namely how to endow protein sequences, upon translation from the corresponding genes, with structural and functional features. Routinely this operation is electronically done by deriving and integrating information from previous knowledge. The reference database for protein sequences is UniProtKB divided into two sections, UniProtKB/TrEMBL which is automatically annotated and not reviewed and UniProtKB/Swiss-Prot which is manually annotated and reviewed. The annotation process is essentially based on sequence similarity search. The question therefore arises as to which extent annotation based on transfer by inheritance is valuable and specifically if it is possible to statistically validate inherited features when little homology exists among the target sequence and its template(s). Results In this paper we address the problem of annotating protein sequences in a statistically validated manner considering as a reference annotation resource UniProtKB. The test case is the set of 48,298 proteins recently released by the Critical Assessment of Function Annotations (CAFA) organization. We show that we can transfer after validation, Gene Ontology (GO) terms of the three main categories and Pfam domains to about 68% and 72% of the sequences, respectively. This is possible after alignment of the CAFA sequences towards BAR+, our annotation resource that allows discriminating among statistically validated and not statistically validated annotation. By comparing with a direct UniProtKB annotation, we find that besides validating annotation of some 78% of the CAFA set, we assign new and statistically validated annotation to 14.8% of the sequences and find new structural templates for about 25% of the chains, half of which share less than 30% sequence identity to the corresponding template/s. Conclusion Inheritance of annotation by transfer generally requires a careful selection of the identity value among the target and the template in order to transfer structural and/or functional features. Here we prove that even distantly remote homologs can be safely endowed with structural templates and GO and/or Pfam terms provided that annotation is done within clusters collecting cluster-related protein sequences and where a statistical validation of the shared structural and functional features is possible.

Summary: The annotation of membrane proteins in proteomes is an important problem of Computational Biology, especially after the development of high-throughput techniques that allow fast and efficient genome sequencing. Among membrane proteins, transmembrane β-barrels (TMBBs) are poorly represented in the database of protein structures (PDB) and difficult to identify with experimental approaches. They are, however, extremely important, playing key roles in several cell functions and bacterial pathogenicity. TMBBs are included in the lipid bilayer with a β-barrel structure and are presently found in the outer membranes of Gram-negative bacteria, mitochondria and chloroplasts. Recently, we developed two top-performing methods based on machine-learning approaches to tackle both the detection of TMBBs in sets of proteins and the prediction of their topology. Here, we present our BETAWARE program that includes both approaches and can run as a standalone program on a linux-based computer to easily address in-home massive protein annotation or filtering. Availability and implementation: http://www.biocomp.unibo.it/∼savojard/betawarecl Contact: piero.fariselli@unibo.it Supplementary information: Supplementary data are available at Bioinformatics online.

Background Various computational methods are presently available to classify whether a protein variation is disease-associated or not. However data derived from recent technological advancements make it feasible to extend the annotation of disease-associated variations in order to include specific phenotypes. Here we tackle the problem of distinguishing between genetic variations associated to cancer and variations associated to other genetic diseases. Results We implement a new method based on Support Vector Machines that takes as input the protein variant and the protein function, as described by its associated Gene Ontology terms. Our approach succeeds in discriminating between germline variants that are likely to be cancer-associated from those that are related to other genetic disorders. The method performs with values of 90% accuracy and 0.61 Matthews correlation coefficient on a set comprising 6478 germline variations (16% are cancer-associated) in 592 proteins. The sensitivity and the specificity on the cancer class are 69% and 66%, respectively. Furthermore the method is capable of correctly excluding some 96% of 3392 somatic cancer-associated variations in 1983 proteins not included in the training/testing set. Conclusions Here we prove feasible that a large set of cancer associated germline protein variations can be successfully discriminated from those associated to other genetic disorders. This is a step further in the process of protein variant annotation. Scoring largely improves when protein function as encoded by Gene Ontology terms is considered, corroborating the role of protein function as a key feature for a correct annotation of its variations.

Motivation: Transmembrane β-barrels (TMBBs) are extremely important proteins that play key roles in several cell functions. They cross the lipid bilayer with β-barrel structures. TMBBs are presently found in the outer membranes of Gram-negative bacteria and of mitochondria and chloroplasts. Loop exposure outside the bacterial cell membranes makes TMBBs important targets for vaccine or drug therapies. In genomes, they are not highly represented and are difficult to identify with experimental approaches. Several computational methods have been developed to discriminate TMBBs from other types of proteins. However, the best performing approaches have a high fraction of false positive predictions. Results: In this article, we introduce a new machine learning approach for TMBB detection based on N-to-1 Extreme Learning Machines that significantly outperforms previous methods achieving a Matthews correlation coefficient of 0.82, a probability of correct prediction of 0.92 and a sensitivity of 0.73. Availability: The method and the cross-validation sets are available at the web page http://betaware.biocomp.unibo.it/BetAware. Contact: piero.fariselli@unibo.it

Correlated mutations in proteins are believed to occur in order to preserve the protein functional folding through evolution. Their values can be deduced from sequence and/or structural alignments and are indicative of residue contacts in the protein three- dimensional structure. A correlation among pairs of residues is routinely evaluated with the Pearson correlation coefficient and the MCLACHLAN similarity matrix. In literature there is no justification for the adoption of the MCLACHLAN instead of other substitution matrices. In this paper, we approach the problem of computing the optimal similarity matrix for contact prediction with correlated mutations, i.e. the the similarity matrix that maximizes the accuracy of contact prediction with correlated mutations. We describe an optimization procedure, based on the gradient descent method, for computing the optimal similarity matrix and perform an extensive number of experimental tests. Our tests show that there is a large number of optimal matrices that perform similarly to MCLACHLAN. We also obtain that the upper limit to the accuracy achievable in protein contact prediction is independent of the optimized similarity matrix. This suggests that the poor scoring of the correlated mutations approach may be due to the choice of the linear correlation function in evaluating correlated mutations.

MemPype is a Python-based pipeline including previously published methods for the prediction of signal peptides (SPEP), glycophosphatidylinositol (GPI) anchors (PredGPI), all-alpha membrane topology (ENSEMBLE), and a recent method (MemLoci) that specifically discriminates the localization of eukaryotic membrane proteins in: 'cell membrane', 'internal membranes', 'organelle membranes'. MemLoci scores with accuracy of 70% and generalized correlation coefficient (GCC) of 0.50 on a rigorous homology-unbiased validation set and overpasses other predictors for subcellular localization. The annotation process is based both on inheritance through homology and computational methods. Each submitted protein first retrieves, when available, up to 25 similar proteins (with sequence identity ≥50% and alignment coverage ≥50% on both sequences). This helps the identification of membrane-associated proteins and detailed localization tags. Each protein is also filtered for the presence of a GPI anchor [0.8% false positive rate (FPR)]. A positive score of GPI anchor prediction labels the sequence as exposed to 'Cell surface'. Concomitantly the sequence is analysed for the presence of a signal peptide and classified with MemLoci into one of three discriminated classes. Finally the sequence is filtered for predicting its putative all-alpha protein membrane topology (FPR

Single residue mutations in proteins are known to affect protein stability and function. As a consequence, they can be disease associated. Available computational methods starting from protein sequence/structure can predict whether a mutated residue is or not disease associated and whether it is promoting instability of the protein-folded structure. However, the relationship among stability changes in proteins and their involvement in human diseases still needs to be fully exploited. Here, we try to rationalize in a nutshell the complexity of the question by generalizing over information already stored in public databases. For each single aminoacid polymorphysm (SAP) type, we derive the probability of being disease-related (Pd) and compute from thermodynamic data three indexes indicating the probability of decreasing (P-), increasing (P+), and perturbing the protein structure stability (Pp). Statistically validated analysis of the different P/Pd correlations indicate that Pd best correlates with Pp. Pp/Pd correlation values are as high as 0.49, and increase up to 0.67 when data variability is taken into consideration. This is indicative of a medium/good correlation among Pd and Pp and corroborates the assumption that protein stability changes can also be disease associated at the proteome level.

Protein sequence annotation is a major challenge in the postgenomic era. Thanks to the availability of complete genomes and proteomes, protein annotation has recently taken invaluable advantage from cross-genome comparisons. In this work, we describe a new non hierarchical clustering procedure characterized by a stringent metric which ensures a reliable transfer of function between related proteins even in the case of multidomain and distantly related proteins. The method takes advantage of the comparative analysis of 599 completely sequenced genomes, both from prokaryotes and eukaryotes, and of a GO and PDB/SCOP mapping over the clusters. A statistical validation of our method demonstrates that our clustering technique captures the essential information shared between homologous and distantly related protein sequences. By this, uncharacterized proteins can be safely annotated by inheriting the annotation of the cluster. We validate our method by blindly annotating other 201 genomes and finally we develop BAR (the Bologna Annotation Resource), a prediction server for protein functional annotation based on a total of 800 genomes (publicly available at http://microserf.biocomp.unibo.it/bar/).

Motivation: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. Results: 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. Availability: The dataset is available at http://www.biocomp.unibo.it/∼lisa/coiled-coils. The predictor is freely available at http://gpcr.biocomp.unibo.it/cgi/predictors/cchmmprof/pred_cchmmprof.cgi. Contact: piero@biocomp.unibo.it

Most of the cellular functions are the result of the concerted action of protein complexes forming pathways and networks. For this reason, efforts were devoted to the study of protein-protein interactions. Large-scale experiments on whole genomes allowed the identification of interacting protein pairs. However residues involved in the interaction are generally not known and the majority of the interactions still lack a structural characterization. A crucial step towards the deciphering of the interaction mechanism of proteins is the recognition of their interacting surfaces, particularly in those structures for which also the most recent interaction network resources do not contain information. To this purpose, we developed a neural network-based method that is able to characterize protein complexes, by predicting amino acid residues that mediate the interactions. All the Protein Data Bank (PDB) chains, both in the unbound and in the complexed form, are predicted and the results are stored in a database of interaction surfaces (http://gpcr.biocomp.unibo.it/zenpatches). Finally, we performed a survey on the different computational methods for protein-protein interaction prediction and on their training/testing sets in order to highlight the most informative properties of protein interfaces.

1. Why membrane proteins? 2 Many functions 3 Bioinformatics and membrane proteins: is it feasible to predict the 3D structure of a membrane protein? 4 Predicting the topology of membrane proteins 5 How many methods to predict membrane protein topology? 5.1 From theory to practice 6 Benchmarking the predictors of transmembrane topology 6.1 Testing on membrane proteins of known structure and topology 6.2 Topological experimental data 6.3 Validation towards experimental data 7 How many membrane proteins in the Human genome? 8 Membrane proteins and genetic diseases: PhD-SNP at work 9 Last but not least: 3D MODELLING of membrane proteins 10 What can currently be done in practice? 11 Can we improve?

Structure prediction of proteins is a difficult task as well as prediction of protein-protein interaction. When no homologous sequence with known structure is available for the target protein, search of distantly related proteins to the target may be done automatically (fold recognition/threading). However, there are difficult proteins for which still modeling on the basis of a putative scaffold is nearly impossible. In the following, we describe that for some specific examples, human expertise was able to derive alignments to proteins of similar function with the aid of machine learning-based methods specifically suited for predicting structural features. The manually curate search of putative templates was successful in generating low-resolution three-dimensional (3D) models in at least two cases: the human tissue transglutaminase and the alcohol dehydrogenase from Sulfolobus solfataricus. This is based on the structural comparison of the model with the 3D protein structure that became available after prediction. For protein-protein interaction, a knowledge-based method can give predictions of putative interaction patches on the protein surface; this feature may help in adding additional weight to specific nodes in nets of interacting proteins.

Compartmentalization plays a major role in eukaryotic cells by making possible the fine regulation of complex biochemical pathways. Each protein needs the right biochemical context to operate, therefore the knowledge of the subcellular localization of a protein is essential in order to understand its functions and its pattern of interactions in protein networks. BaCelLo is a predictor for the subcellular localization of eukaryotic proteins and it is based on several Support Vector Machines (SVMs) arranged in a decision tree (Fig 1). Starting from the residue sequence, BaCelLo discriminates five different localizations: secretory pathway, cytoplasm, nucleus, mitochondrion and chloroplast. The predictor analyzes the protein residue sequence and its evolutionary profile considering information from the whole sequence and from its N- and C-terminal regions. Three different predictors are available for three different eukaryotic kingdoms: Metazoa, Viridiplantae and Fungi. The distinctive features of BaCelLo are: 1. a homology-reduced dataset for training and testing the predictor, in order to avoid redundancy. This dataset was compiled starting from the Swissprot data base (release 48) and contains proteins whose subcellular localization was experimentally annotated. The dataset was reduced by similarity so that no protein in the dataset share more than 30% identity; 2. the implementation of three kingdom-specific predictors to take into account differences in subcellular localization mechanisms; 3. the evolutionary profile to extract evolutionary information from the residue sequence. 4. a hierarchic tree for the predictions; 5. the introduction of a unique balancing procedure in SVMs that corrects the biases between the different classes due to the disproportions in the training set . BaCelLo proved to outperform all the other state-of-art methods publicly available, when validated on a set of protein sequences independent of the training set.

Background Structure prediction of membrane proteins is still a challenging computational problem. Hidden Markov models (HMM) have been successfully applied to the problem of predicting membrane protein topology. In a predictive task, the HMM is endowed with a decoding algorithm in order to assign the most probable state path, and in turn the labels, to an unknown sequence. The Viterbi and the posterior decoding algorithms are the most common. The former is very efficient when one path dominates, while the latter, even though does not guarantee to preserve the HMM grammar, is more effective when several concurring paths have similar probabilities. A third good alternative is 1-best, which was shown to perform equal or better than Viterbi. Results In this paper we introduce the posterior-Viterbi (PV) a new decoding which combines the posterior and Viterbi algorithms. PV is a two step process: first the posterior probability of each state is computed and then the best posterior allowed path through the model is evaluated by a Viterbi algorithm. Conclusion We show that PV decoding performs better than other algorithms when tested on the problem of the prediction of the topology of beta-barrel membrane proteins.

Motivation: The prediction of protein stability change upon mutations is key to understanding protein folding and misfolding. At present, methods are available to predict stability changes only when the atomic structure of the protein is available. Methods addressing the same task starting from the protein sequence are, however, necessary in order to complete genome annotation, especially in relation to single nucleotide polymorphisms (SNPs) and related diseases. Results: We develop a method based on support vector machines that, starting from the protein sequence, predicts the sign and the value of free energy stability change upon single point mutation. We show that the accuracy of our predictor is as high as 77% in the specific task of predicting the ΔΔG sign related to the corresponding protein stability. When predicting the ΔΔG values, a satisfactory correlation agreement with the experimental data is also found. As a final blind benchmark, the predictor is applied to proteins with a set of disease-related SNPs, for which thermodynamic data are also known. We found that our predictions corroborate the view that disease-related mutations correspond to a decrease in protein stability. Availability: http://gpcr2.biocomp.unibo.it/cgi/predictors/I-Mutant2.0/I-Mutant2.0.cgi Contact: casadio@alma.unibo.it

Detection of homologous proteins with low-sequence identity to a given target (remote homologues) is routinely performed with alignment algorithms that take advantage of sequence profile. In this article, we investigate the efficacy of differ- ent alignment procedures for the task at hand on a set of 185 protein pairs with similar structures but low-sequence similarity. Criteria based on the SCOP label detection and MaxSub scores are adopted to score the results. We investigate the efficacy of alignments based on sequence-sequence, sequence- profile, and profile-profile information. We confirm that with profile-profile alignments the results are better than with other procedures. In addition, we report, and this is novel, that the selection of the results of the profile-profile alignments can be im- proved by using Shannon entropy, indicating that this parameter is important to recognize good pro- file-profile alignments among a plethora of meaning- less pairs. By this, we enhance the global search accuracy without losing sensitivity and filter out most of the erroneous alignments. We also show that when the entropy filtering is adopted, the quality of the resulting alignments is comparable to that com- puted for the target and template structures with CE, a structural alignment program. Proteins 2004

MOTIVATION: ConSeq is a web server for the identification of biologically important residues in protein sequences. Functionally important residues that take part, e.g. in ligand binding and protein-protein interactions, are often evolutionarily conserved and are most likely to be solvent-accessible, whereas conserved residues within the protein core most probably have an important structural role in maintaining the protein's fold. Thus, estimated evolutionary rates, as well as relative solvent accessibility predictions, are assigned to each amino acid in the sequence; both are subsequently used to indicate residues that have potential structural or functional importance. AVAILABILITY: The ConSeq web server is available at http://consMOTIVATION: ConSeq is a web server for the identification of biologically important residues in protein sequences. Functionally important residues that take part, e.g. in ligand binding and protein-protein interactions, are often evolutionarily conserved and are most likely to be solvent-accessible, whereas conserved residues within the protein core most probably have an important structural role in maintaining the protein's fold. Thus, estimated evolutionary rates, as well as relative solvent accessibility predictions, are assigned to each amino acid in the sequence; both are subsequently used to indicate residues that have potential structural or functional importance. AVAILABILITY: The ConSeq web server is available at http://conseq.bioinfo.tau.ac.il/ SUPPLEMENTARY INFORMATION: The ConSeq methodology, a description of its performance in a set of five well-documented proteins, a comparison to other methods, and the outcome of its application to a set of 111 proteins of unknown function, are presented at http://conseq.bioinfo.tau.ac.il/ under 'OVERVIEW', 'VALIDATION', 'COMPARISON' and 'PREDICTIONS', respectively.eq.bioinfo.tau.ac.il/ SUPPLEMENTARY INFORMATION: The ConSeq methodology, a description of its performance in a set of five well-documented proteins, a comparison to other methods, and the outcome of its application to a set of 111 proteins of unknown function, are presented at http://conseq.bioinfo.tau.ac.il/ under 'OVERVIEW', 'VALIDATION', 'COMPARISON' and 'PREDICTIONS', respectively.

Background Peptidases are proteolytic enzymes responsible for fundamental cellular activities in all organisms. Apparently about 2–5% of the genes encode for peptidases, irrespectively of the organism source. The basic peptidase function is "protein digestion" and this can be potentially dangerous in living organisms when it is not strictly controlled by specific inhibitors. In genome annotation a basic question is to predict gene function. Here we describe a computational approach that can filter peptidases and their inhibitors out of a given proteome. Furthermore and as an added value to MEROPS, a specific database for peptidases already available in the public domain, our method can predict whether a pair of peptidase/inhibitor can interact, eventually listing all possible predicted ligands (peptidases and/or inhibitors). Results We show that by adopting a decision-tree approach the accuracy of PROSITE and HMMER in detecting separately the four major peptidase types (Serine, Aspartic, Cysteine and Metallo- Peptidase) and their inhibitors among a non redundant set of globular proteins can be improved by some percentage points with respect to that obtained with each method separately. More importantly, our method can then predict pairs of peptidases and interacting inhibitors, scoring a joint global accuracy of 99% with coverage for the positive cases (peptidase/inhibitor) close to 100% and a correlation coefficient of 0.91%. In this task the decision-tree approach outperforms the single methods. Conclusion The decision-tree can reliably classify protein sequences as peptidases or inhibitors, belonging to a certain class, and can provide a comprehensive list of possible interacting pairs of peptidase/inhibitor. This information can help the design of experiments to detect interacting peptidase/inhibitor complexes and can speed up the selection of possible interacting candidates, without searching for them separately and manually combining the obtained results. A web server specifically developed for annotating peptidases and their inhibitors (HIPPIE) is available at http://gpcr.biocomp.unibo.it/cgi/predictors/hippie/pred_hippie.cgi