An Approach to Parallel Data Mining for Pharmacophore Discovery
Rapid and efficient design of new drugs is a key challenge for the medical industry. Current drug design methodologies can take several years just in the initial chemical evaluation stages before compounds can be created for animal and human testing. This paper presents some recent research results in a new parallel machine learning approach that can expedite the drug design cycle. This learning system is based on data partitioning and parallel data mining by a loosely coupled collection of inductive algorithms. The system has been tested on an eight node Beowulf cluster and an I B M supercomputer comprised of 112 processors to discover new pharmacophores for angiotensinconverting enzyme (A C E) inhibitors. The system shows a linear speedup with the number of processor. A new potential ACE inhibitor pharmacophore with 5 active sites has been discovered during the testing of this parallel system.
