Determination of Specificity Residues in Two Component Systems using Graphlets
This work presents a novel method for the identification of specificity residues in two component systems based on the discovery of graphlet signatures. We use network representations of 3-D structures and sequence of proteins, experimental data and graph-based learning to detect graphlet signatures that potentially are responsible for phosphotranfer specificity between Histidine Kinase (HK)and Response Regulator (RR) domains. This approach is applied to the system of HK and RR in E. coli. Structural regions were found for Histidine Kinases RstB and Response Regulator RstA and confirmed using experimental data.In addition, some hypothetical regions of specificity were proposed to explain cross talk between the HKs that phosphorylate YhfA and the RRs that interact with UhpB. Suchan approach offers the ability to identify domain specificity residues in two component systems in silico.