The main purpose of virtual screening is to rank order hit compounds according to their affinity to the protein target. Several functions and techniques have been developed to achieve this purpose; however, simple and forward methods were found less efficient than the complex techniques. In this study, we aim at establishing a workflow that reduces the shortfalls of single methods by integrating structure- and ligand-based screening, especially when an imbalanced dataset is presented. QSAR predictions, pharmacophoric mapping, and shape similarity add up to the docking scores that are all fused to form a parallel Z-score fusion of the scores. The top-ranked structures in the four screening techniques, in addition to the consensus and molecular mechanics (MMGBSA) rescoring methods were compared with co-crystalized ligands and TDP1 inhibitors in terms of their molecular dynamic simulations, binding free energy, and poses. Interestingly, consensus and MMGBSA rescoring methods prioritized the most chemically-diverse structures among other screening ways. Screening by docking has brought structures with binding modes similar to the co-crystalized ligands, whereas, MMGBSA, consensus, and docking resulted in a dynamic simulation similar to and/or better than the co-crystalized ligands. Besides, QSAR-selected structures showed similar binding modes to the commercial TDP1 inhibitors. Notably, a strong correlation was established between the inhibitory concentrations and marketed TDP1 inhibitors’ binding energy values, thus, the inhibitory potencies of the top-ranked structures in such methods are expected to be in the same molar range. In conclusion, consensus scoring can be taken as an adaptable mainstay virtual screening methodology.