This phenomenon is known as “Erooms law”; the number of new approved drugs remains constant as the resources spent on drug development increase exponentially. Getting out of this rut requires thinking out of the box: a wide range of novel strategies based on a broader paradigm. We have applied our combined experimental-computational toolkit to a variety of such novel approaches. This includes the study of more subtle mechanisms for drug-protein interactions than simply matching the form of the drug molecule to a specific active site. For example, we use computational modeling to assist the development of immunotherapy vectors and cell penetrating peptides and peptide derivatives as antibacterial agents and investigate the role of the lipid membrane in substrate selection of weakly membrane associated proteins. Links to our publications in this area are below.