Abstract: One approach to rapidly predict the functions of an entire proteome is to utilize genomic database information and prediction software. Thus, we used a set of prediction algorithms to computationally define a potential secretome. In Xylella fastidiosa, the N-terminal signal peptide was predicted by SignalP version 2.0. Two hundred and four ORFs were predicted to encode proteins with N-terminal signal peptides. Among these, 173 were predicted with no transmembrane domain (or a single transmembrane domain at the extreme N-terminus). TargetP was used to eliminate proteins with 19 mitochondrial targeting signals, and 154 was predicted not to be GPI-anchored. The final computationally-predicted Xylella fastidiosa secretome was estimated to consist of up to 149 ORFs. One hundred and fourteen secretary signal peptides and 32 twin-arginine signal peptides which occupied 76.5% and 21.5% of secretome respectively, and 3 lipoprotein signal peptides only occupied 2%, preplan-like, bacteriocin and pheromone signal peptides were not found.

Key words: Xylella fastidiosa, secreted protein, signal peptide