Linking a drug’s mechanism of action to a disease-related phenotype is the greatest challenge in pharmaceutical research. Target-based and phenotype-based screening are the two basic tools to identifying drug candidates which display efficiency in triggering disease-related phenotype through a known Mechanism of Action (MoA) via a predefined primary target. In this paper, we present a reverse engineering approach which identify drug candidates from active compounds in phenotypic-based assay while bypassing the biochemical assay on the primary target. This important information is deduced implicitly by taking a reference drug with known MoA, i.e., a known primary target as a template. An in silico protocol is developed to preferentially select from phenotypic-active compounds, those that share the same target profile (primary and secondary target) as the reference drug. This is realized by requiring the same pharmacophore pattern as the reference drug while staying within its molecular envelop. These two constraints ensure a similar action on the primary target, while limiting the risk of digression with respect to the profile of secondary targets of the reference drug. Target-based screening assay can be used later to validate this in silico-based drug candidate selection. This reverse engineering approach can be used to crosslink two therapeutic indications if the reference drug has itself been validated experimentally on both. In that case, we use compound activity data from a phenotypic-based high-throughput screening campaign carried out on the target indication, in our case: malaria. The primary target is chosen indirectly through a reference drug and validated in the first indication (here cancer). If we choose a reference drug which is also validated in the malaria field, this implies that the heterologous version of the primary target in cancer field is also a key player of parasite proliferation within humans. We are using BIX-01294, an inhibitor of human histone lysine methyltransferase (HKMT), as a reference drug against cancer which has been validated experimentally to prevent plasmodium proliferation with human red blood cells through the inhibition of HKMT. The developed in silico protocol extracts general cytotoxic compounds with innovative chemical classes, yet with similar MoA as BIX-01294, i.e., inhibiting the original human HKMT, the biochemical bioassay on the plasmodium version of HKMT being not available in a kit-format. This reverse engineering approach is well adapted to transfer the knowledge associated with drug-targets in one therapeutic area (e.g., cancer) to another therapeutic indication for which the target-based approach is way more difficult. The reference drug plays therefore the role of a chemical probe to investigate targets in this target therapeutic area.