A principal problem that the drug discovery research field confronts is to identify small molecules, modulators of protein function, that are likely to be therapeutically useful. Common practices rely on the screening of vast libraries of small molecules (often 1-2 million molecules) in order to identify a molecule that specifically inhibits or activates the protein function, known as a Lead Molecule. Such a molecule interacts with the required target, but generally lacks the other essential attributes required for a drug candidate (ADME properties). In this project we address the problem of building the optimal lead molecule by developing a multi-objective optimization procedure based on nature-inspired computation and statistical predictive models for high dimensional spaces. The challenging task is to discover the optimal lead molecule through testing only an extremely small number of candidate molecules.