Fig. 1. The Ohm workflow.
The Ohm workflow begins with input of the 3D structure of a protein. Next (moving clockwise), a contact matrix is calculated and a perturbation propagation probability matrix is generated according to Eqs. (1)–(3) (“Methods”). The perturbation propagation probability matrix provides the foundation of the perturbation propagation algorithm. If the position of the active site is known, the perturbation propagation algorithm calculates ACIs of all residues relative to the active site. We further devised a program to identify allosteric hotspots based on ACI values. If the position of both the active site and the allosteric site are known, the perturbation propagation algorithm can determine the allosteric pathways that connect the active site and the allosteric site. Further, Ohm identifies critical residues in allosteric pathways based on Eq. (4) (“Methods”). If neither the active site nor the allosteric site is known, the perturbation propagation algorithm can calculate allosteric correlations of each pair of residues. These correlations are then clustered to interrogate how residues in the protein are allosterically coupled.
