An energetics database to model protein-protein interactions

The human genome contains approximately 2,300 single‐pass membrane proteins, and any two proteins can form a dimer with millions of different conformations. Our understanding of the stability and energetics of these membrane proteins informs the study of various signaling complexes, receptor tyrosine kinases, and immunological complexes. This project creates a tool that uses pre‐computed van der Waals and hydrogen bond interaction data for various amino acid combinations in an array of orientations to model the optimal conformation for any pair of sequences. Our lab showed when two proteins are identical, forming a homodimer, their interaction is stabilized primarily by hydrogen bonds and van der Waals forces. Our current algorithm can accurately predict the structure of symmetric homodimers, but in cases when the two protein sequences differ, additional geometric variation must be considered. This project expands our capabilities to include interactions with two different sequences and asymmetric orientations. This will help elucidate the functionality of uncharacterized membrane proteins and predict the likelihood that a helix interacts with other proteins in the membrane. The algorithm will also estimate the stability of different orientations of the same pair of helices and quantify the likelihood that the two proteins would dimerize. Due to the physical limitations of observing membrane proteins, computational methods and computer simulations improve the resources available for studying signaling interactions in membrane proteins and supports in vitro experiments on these proteins. This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .