Identification of synergistic drug combinations

There are 3492 approved drugs in ChEMBL33 (latest version), resulting in 6M pair-wise unique combinations, making it difficult and costly to experimentally test such a huge number of combinations on animals or clinical studies. Therefore, we have recently started developing computational workflow which can identify potential synergistic drug combinations which are also safe to use. One of our computational workflows is based on drug-target information, chemical structural analysis, gene-mania based network analysis, pathway enrichment analysis to first identify synergistic combinations and then utilized in-house AI based adverse drug reaction method to assess the safety profiles of the shortlisted combinations. Unsafe combinations are filtered out before generating the final output. We are not disclosing yet the details of this method as it is still under development and unpublished.