The second keynote speaker of NCWIT is Lydia Kavraki professor of Computer Science and Bioengineering at Rice University, addressing the topic of computing from a multidisciplinary perspective.
Kavraki is speaking about the intersection of robots and biomulecules. Her group has worked on robots and the motion planning problem. Robots that are most complex need to consider the number of movements (degree of freedoms) they need to person amd the configuration of the space and obstacles they need to operate in. The motion planning is done by a probabilistic RoadMap.
Kavraki makes the connection between her work in robotics to biomolecular science. The understanding of small molecules and how they behave play a key role in drug discovery such as those for the HIV virus. Her vision is that computers can develop a system for drug discovery “in silicon,” therefore significantly reducing the cost and time constraints of the process of drug discovery. The key intervention of computer science in that process is in the fast selection of drug design. She and her group inspired their work on the problem from their robotics work. She applied the ideas of the movements of robots to molecular behavior. Molecules are in constant motion, making them a moving target of extreme complexity, far outpacing the complexity involved in robotic movement. The lessons from robotics to the biomolecular problem involves reducing the parameters of the target to isolate essential motion, and use the sampling methods used in robotics to predict the potential shapes of the molecule. The robotics approach enables her group to significantly reduce the complexity of the molecule, capturing 90% of the motion.