In continuation of the previous reports on a combination of 3D‐quantitative structure–activity relationships (QSAR) with computational molecular dynamics (MD) studies, a new variation of 3D‐QSAR/MD method has been employed for drug‐design as an alternative or supplementary for the typical experimental methods. The presented method is more cost‐effective and less time‐consuming than the previous methods and avoids several restrictions of experimental methods, such as validity estimation, and predictability. For this purpose, seven inhibitors for bromodomain (BRD)‐containing protein, as an important protein in the development of different types of cancer and responsible for oncogenic rearrangements, have been selected to study of their interactions by docking and MD simulations using molecular mechanics/generalized born surface area (MM/GBSA) method. To build the proposed model, a common variant of 3D‐QSAR methods, comparative molecular field analysis has been employed using a dataset of 100 MD‐extracted ligand conformations and their corresponding MM/GBSA BRD4‐binding energies. The results showed excellent predictability of the generated model for both the training set and test groups. Finally, two new inhibitors were selected among total 4000 designed derivatives (generated through evolutionary techniques) using the proposed 3D‐QSAR‐MD model. The potentials of these inhibitors were assessed by MD simulations, which showed the higher inhibitory of these compounds than the previous inhibitors. Therefore, this method showed high potentials for acceleration of the procedure of drug design and a basis for joining researchers in computational biology and pharmaceutical sciences.