Decoupling signal recognition from sequence models of protein secretion.

This paper presents a translocation probability function which models the protein SRP-recruitment process–the first stage of the secretory pathway. The prediction accuracy of a simple sequence model (based on Conditional Random Field) incorporating this function improves substantially compared to a model based on signal peptide cleavage site features. For the discrimination of secretory and non-secretory pathway proteins a correlation coefficient of 0.97 is achieved–assisting efforts to uncover localization and function for the growing numbers of protein sequence data. The sequence model can be used to predict signal peptide cleavage sites with an accuracy of 70%. [ABSTRACT FROM AUTHOR]