QUANTITATIVE STRUCTURE-ACTIVITY RELATIONSHIP OF FLAVONOID P56(LCK) PROTEIN-TYROSINE KINASE INHIBITORS - A NEURAL-NETWORK APPROACH

Specific inhibitors of protein tyrosine kinase as antiproliferative ag ents are instrumental in several aspects of neoplastic disease and hav e found wide interest as potential pharmacological agents, We have app lied an artificial neural network based on a counterpropagation algori thm to develop quantitative structure-activity relationships in a larg e dataset of 105 flavonoid derivatives that inhibit the enzyme p56(lck ) protein tyrosine kinase. The results of such approach were compared with the linear multiregression analysis with regard to the ability to fit biological activity surfaces, predict activity, and explore the n onlinear aspects of the dependence of activity on properties. Excellen t correlation was obtained for both classical and quantum chemical des criptors, and relevance of the descriptors to binding properties of th e enzyme receptor active site is hypothesized.