Small molecule kinase inhibitors comprise a growing class of drugs used in cancer treatment for over decade. As opposed to traditional cancer therapies which interfere with DNA synthesis and repair system, kinase inhibitors are directly aimed in specific molecular targets involved in oncogenesis. Due to their selective activity they entail significantly different toxicity profile from traditional cancer drugs.
One of the recently investigated potential targeted therapies are JAK2 kinase inhibitors which constitute promising treatment strategy in several mieloproliferative disorders. Considerable number of patients suffering from primary mielofibrosis, polycythemia vera or essential thrombocytopenia carry a point mutation in JAK2 gene [V617F] which is assumed to be an oncogenic driver in these malignancies. There are several JAK2 inhibitors under development, however, most of them, including Ruxolitinib, are accompanied with severe hematological side effects. Therefore there is still a strong need for designing more selective compounds with less off-target activity. To fulfill these requirements, in silico methods are employed at early stages of drug development process. Among those are virtual libraries screening and de novo ligand design. Further steps include improving compound activity by QSAR and machine learning analysis, selectivity profiling and ADMET prediction.
Using this approach we have designed and developed a novel JAK2 kinase inhibitor with high activity and excellent selectivity comparing to leading compounds in this class. We present subsequent stages of that process, from idea, through virtual screening to hit profiling and lead molecule.