Abstract

Among the proteins found in Prokaryotes, Transmembrane Beta-Barrels (TMBBs) are particularly relevant, since they play key roles in several cell functions. As their name indicates, they cross the lipid bilayer with β-barrel structures. TMB Bs are presently found in the outer membranes of Gram-negative bacteria and of mitochondria and chloroplasts. Loop exposure outside the bacterial cell membranes makes TMB Bs important targets for vaccine or drug therapies. In genomes, they are not highly represented and are difficult to identify with experimental approaches. Actually, the classical methods have a very high rate of false positive assignments. Here we present two methods that have been developed to discriminate TMB Bs from other types of proteins and to predict the TMBB topology. TMBB detection is 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 pre-diction of 0.92 and a sensitivity of 0.73. For the topology prediction, we introduce a Grammatical-Restrained Hidden Conditional Random Fields (GRHCRFs) as an extension of Hidden Conditional Random Fields (HCRFs). GRHCR Fs while preserving the discriminative character of HCR Fs, can assign labels in agreement with the production rules of a defined grammar. We show that in the task of predicting TMBB topology GRHCR Fs perform better than CRF and HMM models of the same complexity.

References: – Fariselli P., Savojardo C., Martelli P.L., Casadio R., “Grammatical-Restrained Hidden Conditional Random Fields for Bioinformatics Applications”, Algorithms for Molecular Biology 2009, 4:13.

- Savojardo C., Fariselli P., Casadio R., “Improving the detection of transmembrane β-barrel chains with N-to-1 Extreme Learning Machines”, Bioinformatics (2011) 27 (22): 3123-3128.