A Unified Explanation for Drug Repurposing and Pharmacological Pleiotropy Based on Classical and Statistical Thermodynamics.

Drug repurposing is an authentic, emerging, and growing aspect of drug development when the demand for new therapeutic solutions is high. Many repurposed drugs have been discovered by serendipity or a non-ordered process driven by chance and sharp observation. These discoveries provide strong evidence for the existence of pharmacological pleiotropy, a highly ordered process well described by thermodynamics. Pleiotropy is an efficient way of propagating information and maintaining the specificity of a biological message and has been a cornerstone in genetics research over decades. While the definition, scale, diversity, and complexity associated with drug repurposing are well documented, pharmaceutical pleiotropy that is fundamental to our understanding of drug repurposing remains less explored. In this review, we examine pharmacological pleiotropy and its underpinning thermodynamics in drug repurposing. Additionally, we have drawn upon the universality of thermodynamics to provide insights into pharmaceutical pleiotropy. We suggest that, in serendipitous drug discovery, information in the repurposed drug often exceeds what was thought available with the rational design of the drug. Our interest in repurposing is on leveraging this information and knowledge generally once a therapeutic benefit from a new chemical entity (NCE) has been demonstrated. This requires a different process from standard drug discovery, and this repurposed pathway is the focus of our manuscript. In this review, we propose that drug repurposing can be defined using Information theory (Shannon entropy), Boltzmann statistical entropy, and the thermodynamic principles for spontaneity described by Gibbs free energy of binding. We conclude that therapeutics including repurposed drugs are facilitators of information and instructional transfer and that the distinguishing features of pharmacology, Information theory, and statistical mechanics are intimately linked. With advances in artificial intelligence and machine learning, with their strong links to Information theory and statistical mechanics, now is an appropriate time to further explore these relationships.