In the last decade, the development and use of new methods in combinatorial chemistry and high-throughput screening has dramatically increased the number of known biologically active compounds. Paradoxically, the number of drugs reaching the market has not followed the same trend, often because many of the candidate drugs present poor qualities in absorption, distribution, metabolism, excretion, and toxicological properties (ADME-Tox). The ability to recognize and discard bad candidates early in the drug discovery steps would save lost investments in time and money. Machine learning techniques could provide solutions to this problem.
The goal of my research is to develop classifiers that accurately discriminate between active and inactive... |