Estudo computacional da interação de inibidores com quinases dependentes de ciclina

Abstract

Cyclin-dependent kinases (CDKs) comprise an interesting biological system for development of docking protocols and scoring functions, due to the abundance of complexed structures for which binding affinity data is available. Here, we report application of an integrated computational approach to carry out docking against a data set composed of 176 structures of CDK in complex with inhibitors. To our knowledge, this is the largest data set of CDK crystallographic structures submitted to molecular docking simulation. Our results indicate that the proposed strategy for docking against CDKs generates poses with docking root-mean square deviation below 2.0 Å for most of the structures in the data set. In addition, we describe the development of scoring functions tailored to CDKs. Statistical analysis of pre-docking and re-docking results, using the proposed scoring functions for CDKs, indicates that these functions are able to predict affinity with better performance when compared with previously reported benchmarks for CDKs.