Abstract: A system of generating a suitability and/or toxicity score for a compound of interest is described. The system may be trained by obtaining physical properties of compounds; retrieving or predicting physiological binding sites or targets of the compound; and using clinical data for known compounds and interactions. Suitability or toxicity for a compound of interest may be obtained based upon a machine-learning prediction. In such a system, the clinical descriptions for the compound may include at least one of an indication for use of the compound and a therapeutic area of the compound, and also may include the use of gene expression strength metrics.