Survival of cells and maintenance of genome depend on detection and repair of damaged DNA through intricate mechanisms. Cancer treatment relies on chemotherapy or radiation therapy that kills neoplastic cells by causing immense damage to the DNA. In many cases, escalated DNA repair mechanism leads to resistance against these therapies and therefore, there is a need to expand the interest in developing drugs that can sensitize the cells to such therapies by interfering with the DNA repair mechanism. Several studies have suggested a link between over expression of the primary mammalian enzyme, Apurinic/Apyrimidinic Endonuclease (APE1), responsible for abasic (or AP) site removal in the DNA and resistance of these cells to cancer therapy, whereas APE1 down-regulation sensitizes the cells to DNA damaging agents. Thus, the current treatment efficacy can be improved by aiding to selective sensitization of cancer cells and protection of normal cells. In the present study, we have used machine learning based approach by selecting assorted compounds with known activity for APE1 and constructed a range of in silico predictive classification models to discriminate between the inhibitors and non-inhibitors. These models can be applied to numerous other unscreened compounds to select the ones which are more likely to be the inhibitors for APE1. We have further found the common molecular substructures which were associated with the molecular activity of the compounds using a substructure search approach.