Sequence based methods for the prediction and analysis of the structural topology of transmembrane beta barrel proteins

Transmembrane proteins play a major role in the normal functioning of the cell. Many transmembrane proteins act as a drug target and hence are of utmost importance to the pharmaceutical industry. In spite of the significance of transmembrane proteins, relatively few transmembrane 3D structures are available due to experimental bottlenecks. Due to this, it is imperative to develop novel computational methods to elucidate the structure and function of these proteins. The two major classes of transmembrane proteins are helical membrane proteins and transmembrane beta barrel proteins. Relatively more 3D structures of helical membrane proteins have been experimentally determined and in general, the majority of computational methods in the realm of transmembrane proteins deal with helical membrane proteins. However, in the recent years there has been an increased interest in the development of computational methods for the transmembrane beta barrel proteins. In this study, I focus on the transmembrane beta barrel proteins. More specifically, I present here computational methods for the prediction of the exposure status of the residues in the membrane spanning region of the transmembrane beta barrel proteins. To the best of our knowledge, the exposure status prediction is a novel problem in the realm of transmembrane beta barrel proteins. The knowledge about the exposure status of the membrane spanning residues is then used to analyse the structural properties of transmembrane beta strands. The exposure status information is also employed to identify relevant physico-chemical properties that are statistically significantly different in the transmembrane beta strands at the oligomeric interfaces and the rest of the protein surface. A method for the prediction of the beta strands in the membrane spanning regions of putative transmembrane beta barrel proteins from protein sequence has also been developed. The computational method for strand prediction is novel in the respect that it also gives the exposure status information of the residues predicted to be in the predicted transmembrane beta strands. The two computational methods developed in this study have been made available as web services. In the future, the information about the exposure status of the residues in the transmembrane beta strands can be used to identify putative transmembrane beta barrels from proteomic data. The exposure status prediction can also be extended to predict the pore region of transmembrane beta barrel proteins from sequence, which could in turn be used in the function prediction of putative transmembrane beta barrels.