Rajarshi Guha presents at the final fall 09 Chemical Information Retrieval class at Drexel University on December 3, 2009. The audio for the first 45 seconds is a little off but the rest is fine. Implicit and explicit molecular representations in 1D, 2D and 3D formats are introduced. Approaches to quantifying molecular similarity using fingerprinting are discussed, such as the Tanimoto index. The relevance of these methods to drug design in terms of virtual screening and QSAR is explored. The concept of scaffold hopping is presented as a way of speeding drug discovery. Finally, ways of comparing the similarity of molecule libraries are detailed, stressing the importance of generating diversity.
« Hide
