Deriving Structures for Lead Drug Discovery from Cell-Line Screens
Abstract
New calculational methods have been developed and applied towards the discovery of novel drugs for the treatment of breast cancer. The analysis is based on publically available tumor screening data generated in the National Cancer Institute's anticancer drug screen. The growth inhibitory potency of 122 anticancer agents tested in the NCI screen have been analyzed using methods of singular value decomposition (SVD). The analysis yielded clusters of compounds with similar activity patterns and compared these results with their putative mechanism of action (MOA). Clustering according to each compound's vector of screening activity segregated compounds into two groups, clearly discernible on the basis of pattern similarities in their potency. The first group includes compounds that act as DNA-damaging agents while the second group includes compounds that act as inhibitors of biosynthetic enzymes or mitosis. Additional analysis of an expanded set of tested compounds finds that a significant statistical correlation can be found between compounds that have similar functions and compounds with structural similarities. These results represent the first evidence for a strong correlation between cancer screening data and structure. These results provide a basis for further explorations of relationships between structural modalities of potential anticancer agents and measurements of cellular toxicity.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1999
- Accession Number
- ADA385801
Entities
People
- David G. Corell
- Robert L. Jernigan
Organizations
- National Cancer Institute