Computational Prediction of Amino Acids Governing Protein-Membrane Interaction for the PIP₃ Cell Signaling System

Prediction and characterization of how transiently membrane-bound signaling proteins interact with the cell membrane is important for understanding and controlling cellular signal transduction networks. Existing computational methods rely on approximate descriptions of the components of the system or their interactions, and thus are unable to identify residue- or lipid-specific contributions. Our rotational interaction energy profiling method allows rapid evaluation of an electrostatically optimal orientation of a protein for membrane association, as well as the residues or lipid species responsible for its favorability. This enables prediction of which aspects of the protein-membrane interaction to target experimentally, and thus the development of testable hypotheses, as well as providing efficient seeding of molecular dynamics simulations to further characterize the protein-membrane interaction. We illustrate our method on two proteins of the PIP₃ cell signaling system, PTEN and PI3Kα.