Using the TOPS-MODE approach to fit multi-target QSAR models for tyrosine kinases inhibitors

Abstract

Tyrosine kinases constitute an eligible class of target for novel drug discovery. They resulted often overexpressed and/or deregulated in several cancer diseases. Thus, the development of novel tyrosine kinases inhibitors is of value, as well as the finding of novel cheminformatic tools for their design. Among the different ways to rationally design novel compounds, the Quantitative Structure-Activity Relationship (QSAR) plays a key role. The QSAR approach, in fact, allow the prediction of activity against a number of targets (multi-target QSAR), thus leading to models able to predict not only the activity of a compound, but also its selectivity versus a set of targets. Despite it is well known that tyrosine kinase inhibitors have to show multi-kinases inhibitory potency to be useful in anticancer therapy, only few multi-target computational tools have been developed to help medicinal chemists in the design of novel compounds. Herein we present the development of several multi-target classification QSAR (mtc-QSAR) models useful to assess the activity profile of the tyrosine kinases inhibitors.