PASL: Prediction of the alpha-helix transmembrane by pruning the subcellular location

Abstract

We have developed a software tool, called PASL, which predicts the transmembrane region and its topology by pruning the subcellular location. The main virtues of PASL are that it discriminates the integral proteins of the plasma membrane from the intracellular membranes, and it eliminates the possibility of misrecognition of the signal peptide as a transmembrane region. The transmembrane region prediction algorithm, which is based on the Hidden Markov Model, and the ER signal peptide detection architecture, which is based on neural networks, have been used for the actual implementation of a prototype. This paper mainly describes the prototype and how it works. © Springer-Verlag 2004.