Backward smoothing approach to transmembrane protein structure prediction with stochastic dynamical systems

A backward smoothing approach was used to predict transmembrane protein structure. The proposed scheme utilizes a stochastic dynamical system with two-dimensional vector trajectories consisting of a hydropathy index and formal charge. Given a sequence of amino acids of unknown structure, each residue was predicted to be or not to be in a transmembrane region by a backward smoothing process. The proposed scheme, relative to the standard smoothing schemes applicable to the proposed model, was evaluated using publicly available benchmarking sequences. Our algorithm was found to yield reasonably good results.