TOPTMH: TOPOLOGY PREDICTOR FOR TRANSMEMBRANE α-HELICES

摘要

Alpha-helical transmembrane proteins mediate many key biological processes and repre-sent 20%-30% of all genes in many organisms. Due to the difficulties in experimentallydetermining their high-resolution 3D structure, computational methods to predict thelocation and orientation of transmembrane helix segments using sequence informationare essential. We present TOPTMH, a new transmembrane helix topology predictionmethod that combines support vector machines, hidden Markov models, and a widelyused rule-based scheme. The contribution of this work is the development of a predictionapproach that first uses a binary SVM classifier to predict the helix residues and then itemploys a pair of HMM models that incorporate the SVM predictions and hydropathy-based features to identify the entire transmembrane helix segments by capturing thestructural characteristics of these proteins. TOPTMH outperforms state-of-the-art pre-diction methods and achieves the best performance on an independent static benchmark.