Characterizing Evolutionary and Physical Energy Landscapes of Evolving, Devolving, and Random Sequences*

Due to evolutionary constraints, protein sequences cooperatively fold into unique, energetically minimized three-dimensional structures in timescales as short as microseconds. The evolutionary importance of maintaining a stable protein structure can be quantified by the selection temperature. The selection temperature can be calculated using AWSEM, a coarse-grained free energy function, and DCA, a protein-specific information theoretic Hamiltonian. Expanding upon the list of previously studied proteins, we quantified selection temperatures of multiple proteins with a wide-array of biological functions. We found selection temperatures to be below physiological temperatures, indicating that folding energetics are an important evolutionary consideration. Furthermore, we incorporated temporal information by calculating the energetics of pseudogenes, devolved and randomly mutated protein-coding genes. We found that pseudogenes are more unstable than protein sequences, as a function of their degree of devolution.