Prediction of anti-inflammatory activity spectrum (COX/PGE2) of compounds with anti-cancer activity in colon

Abstract

COX and PGE2 are inflammatory cytokines that affect various colorectal cancer-related pathways. The Discovery of compounds with the capability of targeting (COX/PGE2) related colorectal cancer receptors is one of the treatment development strategies for colorectal cancer. Objective: To predict colorectal cancer-associated targets that could be targeted with the compounds that are inhibitors or antagonists of the COX/PGE2 and its receptor pathway a systematic data-mining method was developed. Methods: 32 targets related to the (COX/PGE2) pathway and associated with colorectal cancer were selected using the KEGG database. The related compounds were extracted from Chembl25 and the Tanimoto similarity index was calculated using Morgan fingerprints. Molecular docking methods were used to investigate the interaction of the selected compounds with the predicted targets. a systematic data mining and the computational method was developed using the KNIME platform. Results: A total of 15900 compounds were extracted for 32 targets, leaving 14800 compounds after the application of rule-based filters. compounds with a similarity index above 0.7, were selected for the final study. Among targets, there was empirical evidence to inhibit the predicted target for PI3 kinase-mTOR, GSK3-CD1, GSK3-AKT, PI3 kinase-AKT pair, while there was no empirical evidence for the predicted target (GSK3-I kappa B kinase). To investigate the probability of the interaction of the compound with predicted target, the interaction of one inhibitor compound for each pair was investigated by molecular docking method in comparison with the available inhibitor compound and the results were discussed. Conclusion: The developed data mining method significantly reduced the time required for similar studies and has made it possible to conduct studies with several hundred targets and several hundred thousand ligands in a total time of 20 minutes. The results of this method can be used for target identification and lead compound finding purposes.