Recent progress in science and medicine has led to extensive advances in genetics, molecular medicine, and drug discovery. Despite these advances, the development of safe and effective drugs is still one of the most complex and expensive endeavors in modern medicine. Fortunately, advances in cheminformatics and and bioinformatics have enabled the development of new methods to identify drugs and quantify the effects of known drugs. Large databases of chemical compounds and their molecular and cellular effects are publicly available and can be combined with computational methods to aid drug discovery and development. Advances in genome sequencing and genomic medicine have led to increased research into pharmacogenetic markers that are predictive of individual patient responses to drugs. These developments provide opportunities to establish connections between features of drugs and the phenotypes induced by those drugs. In this thesis I present methods to quantify the connections between chemical features of drugs at the molecular, cellular, and organismal level. Specifically, I present a method for predicting activities among structurally diverse sets of drugs, a method for predicting up-regulation of genes based on chemical features of drugs, and a quantified analysis of the clinical impacts of genotype guided therapies. The work presented in this thesis provides methods to support drug discovery, drug development, and the establishment of safe and effective therapies.