Heat shock protein 90 (HSP90) and human epidermal growth factor receptor 2 protein (HER2) are involved in several signal pathways for cancer cell proliferation. We focused on these two hallmarkers to design the dual-targeted inhibitors for cancer therapy. The comparative molecular field analysis (CoMFA) and pharmacophore analysis were employed for generating the predictive models. According the leave-one out (LOO) cross-validation, the CoMFA models obtained the significant r2 values of 0.978 and 0.974 for HSP90 and HER2, respectively. The contour maps of both targets indicated that there were amount of similar bulky favors area. Besides, the cost difference of pharmacophore models was 48.539 for 70% correlation with the experiment. The queries at 3-N and 6-N position of purine-based compound had the similar distributions. This study provided important information for design the HER2 and HSP90 dual-targeted inhibitors.
IEEE/ACM Transactions on Computational Biology and Bioinformatics