A Novel Approach to Predict Core Residues on Cancer-Related DNA-Binding Domains

摘要

Protein–DNA interactions are involved in different cancer pathways. In particular, the DNA-binding domains of proteins can determine where and how gene regulatory regions are bound in different cell lines at different stages. Therefore, it is essential to develop a method to predict and locate the core residues on cancer-related DNA-binding domains. In this study, we propose a computational method to predict and locate core residues on DNA-binding domains. In particular, we have selected the cancer-related DNA-binding domains for in-depth studies, namely, winged Helix Turn Helix family, homeodomain family, and basic Helix-Loop-Helix family. The results demonstrate that the proposed method can predict the core residues involved in protein–DNA interactions, as verified by the existing structural data. Given its good performance, various aspects of the method are discussed and explored: for instance, different uses of prediction algorithm, different protein domains, and hotspot threshold setting.