Background: Large-scale sequencing has characterized an enormous number of genetic variations (GVs), and the functional analysis of GVs is fundamental for understanding the difference of disease susceptibility and therapeutic response among populations.Methods: Combined with a sequence-based predictor, known phosphorylation and protein-protein interaction information, we computationally detected 4,490 potential phosSNPs (Phosphorylation-related non-synonymous single nucleotide polymorphisms) including 275 known disease-associated SNPs that dramatically rewire the human phosSNP-associated kinase-substrate network.