SORIN AVRAM, STEFANA AVRAM, CRISTINA DEHELEAN EFFECTIVE PREDICTOR OF HUMAN MAST CELL TRYPTASE INHIBITORS
Mast cells (MCs) play a key role in the immune response to pathogens, in allergic and inflammatory reactions. Recently, increasing evidence has linked degranulated MCs to cancer through the presence of proangiogenic factors. Tryptase, a protease released from activated MCs granules has emerged as a potential target for tumor treatment. In this study, we aimed to develop a random forest model which is able to effectively classify tryptase inhibitors based on two-dimensional pharmacophore fingerprints. The external validation of the predictor demonstrated excellent performance with an AUC value of 0.953. Moreover, we highlight essential variables identified by the algorithm embedded in random forest. The hereby proposed classifier provides new means for the identification and optimization of tryptase inhibitors which are promising anti-angiogenic agents in the treatment of cancers.
Keywords: mast cells, tryptase inhibitor, random forest, prediction model