Welcome to eKPI

Kinase-mediated phosphorylation signaling plays pivotal roles in regulating almost all biological processes. Recent advancement of high-throughput proteomics techniques boosted the identification of massive phosphorylation sites, for which the regulatory kinases is largely (>95%) unknown. Thus, dissecting the specific interconnections between kinases and phosphorylation sites (phosphosites) is urgently needed. Currently, a number of computational resources could infer potential specific regulatory relations (defined as interconnections) between kinases and phosphorylation substrates/sites (KPIs) based on protein sequence/structure features and protein-protein interaction information. However, the false discovery rates are very high in actual application for experiment validation because the negative hits are dozens of times more than the positive ones.

Here we present the eKPI database that integrate and rank the potential of KPIs based on the 108 correlation data matrixes from 23 multi-omics research projects in 22 types of cancers among the quantitative datasets. The qualitative evidences were integrated from 275 research projects. These correlations calculated from omics datasets FIRSTLY provided the direct evidences in 'real world' for the KPIs. In addition, all the experimental validated KPIs and the prediction results from 7 softwares were also integrated for the 200,198 phosphorylation sites.