Protein Structure and Function. Application of Bioinformatics Methods - John Rigden 2014

Membrane Protein Structure Prediction
Prediction of Transmembrane Protein Topology
Genome-Wide Association Studies

Large-scale genome and proteome sequencing projects frequently uncover novel Proteins whose cellular Localization and Functions remain largely unknown. While several of the Methods described above can predict TM protein topology with high accuracy, the arsenal of tools available for distinguishing TM proteins from globular ones is rather limited. Such differentiation requires dedicated methods and standalone software packages, as web-based prediction tools are ill-suited for Processing massive datasets. Several methods suitable for genome-wide analysis of TM alpha-helices and Structure/56.html">Beta-barrels are summarized in Tables 4.3 and 4.4. Overall, the error rate is significantly reduced by incorporating a preliminary filtering step. Because many Globular proteins contain signal Peptides, they are often misclassified as transmembrane proteins during initial analysis. Preliminary filtering utilizing tools such as SignalP and TargetP effectively screens out signal and transit peptides. Currently, state-of-the-art prediction methods achieve error rates of less than 1% for alpha-helical TM proteins (Jones 2007) and less than 6% for beta-barrel TM proteins (Park et al. 2005). Fig. 4.6 illustrates the application of the alpha-helical TM Protein Differentiation method to a set of proteomes.