Display omitted •A two-step protocol enables more effcient computational receptor-ligand docking.•Machine learning-based pipeline provides an effective way to evaluate binding regions for protein–ligand complexes.•The algorithm works for both peptide and non-peptide ligands, and also in cases where the binding site is buried.•When binding region data are not available, the algorithm provides a fast and effective alternative to classical predictors.•The approach can be generalized by extending the training dataset to include other classes of compounds. We present a combined computational approach to protein–ligand binding, which consists of two steps: (1) a deep neural network is used to locate a binding region on a target protein, and (2) molecular docking of a ligand is performed within the specified region to obtain the best pose using Autodock Vina. Our in-house designed neural network was trained using the PepBDB dataset. Although the training dataset consisted of protein-peptide complexes, we show that the approach is not limited to peptides, but also works remarkably well for a large class of non-peptide ligands. The results are compared with those in which the binding region (first step) was provided by Accluster. In cases where no prior experimental data on the binding region are available, our deep neural network provides a fast and effective alternative to classical software for its localization. Our code is available at https://github.com/mksmd/NNforDocking.