From single to multi-target kinases inhibitors: using the old-age kinases inhibitors data to face the new-age

The cell cycle regulatory systems alterations may cause cancer onset and progression and metastasis. Among the proteins involved in signal recognition, transduction and amplification, tyrosine kinases (TKs) plays a key role. In the last years several efforts have been made to discover potent and selective TKs inhibitors. However, it has been recently demonstrated that compounds with high selectivity inhibition profile may undergo to drug resistance onset. On the contrary, compounds with a broad TKs inhibition profile could be more useful for anticancer therapy. All this findings have caused a renewal in the TKIs development approach. The non-serendipitous based discovery of multi-target TKIs thus is a field of major interest, as well as the development of novel approaches to the rational design of compounds with specific inhibitory profile. Herein we present an overview of the chemoinformatic tools developed in our research group to screen and design novel potential TKs inhibitors. Several QSAR models have been developed starting from literature data and have been validated through both external validation sets and novel compounds evaluation. A comparison between classical and modern approaches (i.e. a novel clusterization-based method and use of artificial neural network) are also presented. Finally, the preliminary results based on docking studies will be discussed.