Signal peptide discrimination and cleavage site identification using SVM and NN

Abstract

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues.Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. a b s t r a c tAbout 15% of all proteins in a genome contain a signal peptide (SP) sequence, at the N-terminus, that targets the protein to intracellular secretory pathways. Once the protein is targeted correctly in the cell, the SP is cleaved, releasing the mature protein. Accurate prediction of the presence of these short aminoacid SP chains is crucial for modelling the topology of membrane proteins, since SP sequences can be confused with transmembrane domains due to similar composition of hydrophobic amino acids. This paper presents a cascaded Support Vector Machine (SVM)-Neural Network (NN) classification methodology for SP discrimination and cleavage site identification. The proposed method utilises a dual phase classification approach using SVM as a primary classifier to discriminate SP sequences from Non-SP. The methodology further employs NNs to predict the most suitable cleavage site candidates. In phase one, a SVM classification utilises hydrophobic propensities as a primary feature vector extraction using symmetric sliding window amino-acid sequence analysis for discrimination of SP and Non-SP. In phase two, a NN classification uses asymmetric sliding window sequence analysis for prediction of cleavage site identification.The proposed SVM-NN method was tested using Uni-Prot non-redundant datasets of eukaryotic and prokaryotic proteins with SP and Non-SP N-termini. Computer simulation results demonstrate an overall accuracy of 0.90 for SP and Non-SP discrimination based on Matthews Correlation Coefficient (MCC) tests using SVM. For SP cleavage site prediction, the overall accuracy is 91.5% based on cross-validation tests using the novel SVM-NN model.

Molecular characterization of Vibrio harveyi virulence-associated serine protease and outer membrane protein genes for vaccine development

“…As described by Kazemian et al (2014), the signal peptide is structured as such that at the N-terminus a positively charged n-region is located. The length of residues in the n-region varies from 1 to 12 residues and is followed by the hydrophobic region, 'h-region', of 7 to15 residues.…”

“…It has also been described that signal peptides and N-terminal -barrels regions usually possess a hydrophobic region (Petersen et al;Nielsen et al, 1997). However, transmembrane -helices region have longer hydrophobic regions and do not have cleavage sites compared to signal peptide region that is shorter approximately consists of 7 to 15 residues (Kazemian et al, 2014;Petersen et al, 2011).…”

An ANFIS approach to transmembrane protein prediction

“…Fig. 1 utilizes a prescribed number of moving amino acid residues, in here 15, known as the 'sliding window' to analyze the protein sequences 1jgy_L for the purpose of protein segment identification, where the alpha-helices are continuously found for the middle residue based on the hydrophobic values on both sides of the sliding window [10,16]. The four variables for hydrophobicity and propensity are 'Left_Hydrophobicity', 'Right_Hydrophobicity', 'Left_Propensity' and 'Right_Propensity'.…”

NN approach and its comparison with NN-SVM to beta-barrel prediction

“…SVM has previously been used for the prediction of transmembrane segments to predict the presence of helical segments in transmembrane proteins (Kazemian, White, Palmer-Brown and Yusuf, 2013). The application of SVM in SP topological prediction has largely been focussed on differentiating N-terminal SPs from alphahelical segments (Kazemian, Yusuf and White, 2014) primarily due to the fact that both the segments share a number of traits including their underlying hydrophobic nature.…”

“…To predict the two class representations, SVM scheme maps and constructs a hyperplane in a highdimensional space, which can be used to separate NTM from TM. A good separation generally is attained by the hyperplane that has the biggest margin, since the higher the margin the lower the generalization error of the classifier (Kazemian, Yusuf and White, 2014).…”