Protein phosphorylation prediction: limitations, merits and pitfalls

Protein phosphorylation is a major protein post-translational modification process that plays a pivotal role in numerous cellular processes, such as recognition, signaling or degradation. It can be studied experimentally by various methodologies, including western blot analysis, site-directed mutagenesis, 2D gel electrophoresis, mass spectrometry etc. A number of in silico tools have also been developed in order to predict plausible phosphorylation sites in a given protein. In this review, we conducted a benchmark study including the leading protein phosphorylation prediction software, in an effort to determine which performs best. The first place was taken by GPS 2.2, having predicted all phosphorylation sites with a 83% fidelity while in second place came NetPhos 2.0 with 69%.