Within this project we will try to analyse the specificity of signal transduction in bacterial two-component systems. Therefore, we will use correlated mutations in components of these systems to look at protein:protein interactions specificity determining residues in components of the chemotaxis signalling pathway. More specifically, we will look at those residues which are important in determining the specific interactions of the central histidine protein kinase CheA with its response regulators CheY, CheB and the scaffold protein CheW. In particular, it is known that Rhodobacter sphaeroides has three CheAs, four CheWs, 6 CheYs and two CheBs and that these homologues have differing affinities for each other. By an iterative approach of using bioinformatics to predict specificity determining residues on the basis of multiple genome annotations and subsequent biochemical analysis of these mutations in vitro and in vivo systems using the chemotaxis proteins of Rh. sphaeroides we can develop new and better methodologies for the prediction of protein:protein interactions from genome sequences.