General Departmental Seminar Series
Machine Learning Issues in Computational Drug Design
Christian Lemmen, Senior Research Scientist,
Thursday, Mar 22, 2001, 2:30-3:30 pm
1221 Computer Sciences and Statistics Building
1210 West Dayton St.
Predicting the activity of molecules is a task of key-importance in the drug discovery process. Ever since the fundamental concept that `similar molecules behave similarly' was formulated, people have tried to assess the similarity in order to raise chances for activity. Over the last couple of years we have developed several methods to arrive at 3D-descriptors for the molecules that have proven to be powerful tools in the drug design process.
Simultaneously combinatorial chemistry technology evolved that made an iterative approach of computational and chemical experiments possible. In this presentation we describe this approach, detail on the descriptors, and aim at pointing out important Machine Learning problems to be tackled to make this process even more effective.
Back to General Departmental Seminar Series