The accurate prediction of biomolecular recognition is fundamental to modern drug discovery. Computational chemists seek to optimize interactions between proteins and drug candidates. As nature has optimized these interactions over billions of years, fundamental understanding of evolutionarily conserved interactions is important for the design of selective, high affinity treatments. Herein, several projects are described that leverage such understanding. Identification of conserved residues in drug targets can suggest residues more resistant to mutation, and therefore more attractive for drug discovery. Similarly, analysis of slight differences in protein subpockets can be leveraged to identify regions to target to improve selectivity between related proteins. Examining interactions between functional moieties on the ligand and receptor microenvironments they interact with can identify evolutionarily conserved enthalpic contributions