Searching across hundreds of databases

Our searching services are busy right now. Your search will reload in five seconds.

X
Forgot Password

If you have forgotten your password you can enter your email here and get a temporary password sent to your email.

Evaluation of methods for predicting the topology of beta-barrel outer membrane proteins and a consensus prediction method.

BMC bioinformatics | Jan 12, 2005

BACKGROUND: Prediction of the transmembrane strands and topology of beta-barrel outer membrane proteins is of interest in current bioinformatics research. Several methods have been applied so far for this task, utilizing different algorithmic techniques and a number of freely available predictors exist. The methods can be grossly divided to those based on Hidden Markov Models (HMMs), on Neural Networks (NNs) and on Support Vector Machines (SVMs). In this work, we compare the different available methods for topology prediction of beta-barrel outer membrane proteins. We evaluate their performance on a non-redundant dataset of 20 beta-barrel outer membrane proteins of gram-negative bacteria, with structures known at atomic resolution. Also, we describe, for the first time, an effective way to combine the individual predictors, at will, to a single consensus prediction method. RESULTS: We assess the statistical significance of the performance of each prediction scheme and conclude that Hidden Markov Model based methods, HMM-B2TMR, ProfTMB and PRED-TMBB, are currently the best predictors, according to either the per-residue accuracy, the segments overlap measure (SOV) or the total number of proteins with correctly predicted topologies in the test set. Furthermore, we show that the available predictors perform better when only transmembrane beta-barrel domains are used for prediction, rather than the precursor full-length sequences, even though the HMM-based predictors are not influenced significantly. The consensus prediction method performs significantly better than each individual available predictor, since it increases the accuracy up to 4% regarding SOV and up to 15% in correctly predicted topologies. CONCLUSIONS: The consensus prediction method described in this work, optimizes the predicted topology with a dynamic programming algorithm and is implemented in a web-based application freely available to non-commercial users at http://bioinformatics.biol.uoa.gr/ConBBPRED.

Pubmed ID: 15647112 RIS Download

Mesh terms: Algorithms | Analysis of Variance | Artificial Intelligence | Bacterial Outer Membrane Proteins | Computational Biology | Computer Graphics | Computer Simulation | Data Interpretation, Statistical | Databases, Protein | Evaluation Studies as Topic | Internet | Markov Chains | Membrane Proteins | Models, Chemical | Models, Molecular | Neisseria meningitidis | Neural Networks (Computer) | Programming Languages | Protein Conformation | Protein Structure, Secondary | Protein Structure, Tertiary | Proteins | Reproducibility of Results | Sequence Alignment | Sequence Analysis, Protein | Software | Staphylococcus aureus

Publication data is provided by the National Library of Medicine ® and PubMed ®. Data is retrieved from PubMed ® on a weekly schedule. For terms and conditions see the National Library of Medicine Terms and Conditions.

This is a list of tools and resources that we have found mentioned in this publication.


ConBBPRED

A web tool for the Consensus Prediction of TransMembrane Beta-Barrel Proteins. Prediction of the transmembrane strands and topology of beta-barrel outer membrane proteins is of interest in current bioinformatics research. Several methods have been applied so far for this task, utilizing different algorithmic techniques and a number of freely available predictors exist. The methods can be grossly divided to those based on Hidden Markov Models (HMMs), on Neural Networks (NNs) and on Support Vector Machines (SVMs). In this work, we compare the different available methods for topology prediction of beta-barrel outer membrane proteins. We evaluate their performance on a non-redundant dataset of 20 beta-barrel outer membrane proteins of gram-negative bacteria, with structures known at atomic resolution. Also, we describe, for the first time, an effective way to combine the individual predictors, at will, to a single consensus prediction method. We assess the statistical significance of the performance of each prediction scheme and conclude that Hidden Markov Model based methods, HMM-B2TMR, ProfTMB and PRED-TMBB, are currently the best predictors, according to either the per-residue accuracy, the segments overlap measure (SOV) or the total number of proteins with correctly predicted topologies in the test set. Furthermore, we show that the available predictors perform better when only transmembrane beta-barrel domains are used for prediction, rather than the precursor full-length sequences, even though the HMM-based predictors are not influenced significantly. The consensus prediction method performs significantly better than each individual available predictor, since it increases the accuracy up to 4% regarding SOV and up to 15% in correctly predicted topologies.

tool

View all literature mentions

UCL Bioinformatics Group

Group headed by Professor David Jones, and was originally founded as the Joint Research Council funded Bioinformatics Unit within the Department of Computer Science at University College London. Supports the following tools: Protein Structure Prediction Threading (THREADER) Ab initio folding simulations Secondary structure prediction (PSIPRED) Protein disorder prediction (DISOPRED) Protein domain prediction (DomPred) Database of protein disorder (DisoDB) Protein Sequence Analysis Protein function prediction (ffpred) Metsite: Metal binding residue prediction HSPred : Protein-protein interaction characterisation Amino acid substitution matrices Hidden Markov Models (collaboration with N. Goldman, Cambridge, & J. Thorne, NCSU) Genome Analysis Genomic fold recognition (GenTHREADER) Genome annotation using software agents Protein Structure Classification CATH (collaboration with J. Thornton & C. Orengo, UCL Biochemistry) Transmembrane Protein Modelling MEMSAT & MEMSATSVM Folding In Lipid Membranes (FILM) MEMPACK Biological Applications of Data-mining and Machine Learning Techniques Information extraction for biological research (BioRat) Microarray Analysis Data integration for microarray analysis Data visualization Systems Biology Systems biology applied to stem cells Legacy Services (to be retired shortly) Comparison of structure classifications (CATH/SCOP/FSSP) Genomic Threading Database (GTD)

tool

View all literature mentions

PRED-TMBB

A web tool, based on a Hidden Markov Model, capable of predicting the transmembrane beta-strands of the gram-negative bacteria outer membrane proteins, and of discriminating such proteins from water-soluble ones when screening large datasets. The model is trained in a discriminative manner, aiming at maximizing the probability of the correct prediction rather than the likelihood of the sequences. The training is performed on a non-redundant database consisting of 16 outer membrane proteins (OMP''s) with their structures known at atomic resolution. We show that we can achieve predictions at least as good comparing with other existing methods, using as input only the amino-acid sequence, without the need of evolutionary information included in multiple alignments. The method is also powerful when used for discrimination purposes, as it can discriminate with a high accuracy the outer membrane proteins from water soluble in large datasets, making it a quite reliable solution for screening entire genomes. This web-server can help you run a discriminating process on any amino-acid sequence and thereafter localize the transmembrane strands and find the topology of the loops.

tool

View all literature mentions