Hidden neural networks for transmembrane protein topology prediction

Ioannis A Tamposis; Dimitra Sarantopoulou; Margarita C Theodoropoulou; Evangelia A Stasi; Panagiota I Kontou; Konstantinos D Tsirigos; Pantelis G Bagos

doi:10.1016/j.csbj.2021.11.006

Hidden neural networks for transmembrane protein topology prediction

Comput Struct Biotechnol J. 2021 Nov 8:19:6090-6097. doi: 10.1016/j.csbj.2021.11.006. eCollection 2021.

Authors

Ioannis A Tamposis¹, Dimitra Sarantopoulou^{2

3}, Margarita C Theodoropoulou¹, Evangelia A Stasi¹, Panagiota I Kontou¹, Konstantinos D Tsirigos⁴, Pantelis G Bagos¹

Affiliations

¹ Department of Computer Science and Biomedical Informatics, University of Thessaly, 35100 Lamia, Greece.
² Institute for Translational Medicine and Therapeutics, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
³ Present address: National Institute on Aging, National Institutes of Health, Baltimore, Maryland, USA.
⁴ EMBL-EBI, Wellcome Genome Campus, Cambridge, United Kingdom.

Abstract

Hidden Markov Models (HMMs) are amongst the most successful methods for predicting protein features in biological sequence analysis. However, there are biological problems where the Markovian assumption is not sufficient since the sequence context can provide useful information for prediction purposes. Several extensions of HMMs have appeared in the literature in order to overcome their limitations. We apply here a hybrid method that combines HMMs and Neural Networks (NNs), termed Hidden Neural Networks (HNNs), for biological sequence analysis in a straightforward manner. In this framework, the traditional HMM probability parameters are replaced by NN outputs. As a case study, we focus on the topology prediction of for alpha-helical and beta-barrel membrane proteins. The HNNs show performance gains compared to standard HMMs and the respective predictors outperform the top-scoring methods in the field. The implementation of HNNs can be found in the package JUCHMME, downloadable from http://www.compgen.org/tools/juchmme, https://github.com/pbagos/juchmme. The updated PRED-TMBB2 and HMM-TM prediction servers can be accessed at www.compgen.org.

Keywords: CHMM, Class Hidden Markov Models; CML, Conditional Maximum Likelihood; EM, Expectation-Maximization; HMM, Hidden Markov Models; HNN, Hidden Neural Networks; Hidden Markov Models; Hidden Neural Networks; JUCHMME, Java Utility for Class Hidden Markov Models and Extensions; MCC, Matthews Correlation Coefficient; ML, Maximum Likelihood; MSA, Multiple Sequence Alignment; Membrane proteins; NN, Neural Networks; Neural Networks; Protein structure prediction; SOV, segment overlap; Sequence analysis.