Linear discrimination of transmembrane from non-transmembrane segments in proteins using higher-order crossings

Identification of transmembrane segments in protein sequences is an important issue in the field of bioinformatics. In this study, a method is proposed for linear discrimination between transmembrane and non-transmembrane segments, combining chemical and statistical features of the proteins with higher-order crossings analysis for protein segment classification. The method was tested on human proteins extracted from public available databases and the results have shown a remarkable efficiency of the proposed algorithm to correctly classify the sequence segments under study into two linearly separated classes, for a wide range of transmembrane segment lengths