Increasingly, pharmaceutical industries are taking a more focused approach for high throughput screening (HTS) to find a greater number of interesting drug leads in a more efficient and cost-effective process. Various academic centers are also taking the similar approach to increase their translational research capacity and help find cures sooner for diseases such as cancer. One of the HTS approaches involves computational analysis of a large dataset in order to highlight those compounds most likely to be active in the actual assay, so that a focused subset of compounds can be selected and tested. Computation-based HTS covers a wide range of computer-aided drug design (CADD) technologies - from very fast property predictions to more computational modeling of drug-target binding. This approach would significantly shorten the time needed and dramatically reduce the cost required. We recently established a chemoinformatic laboratory to perform the computation-based HTS and have a database of >1,000,000 small molecule compounds with 3D structure, potential toxicity, ADME and pharmaceutical properties.