Integrative Analysis of Multi-omics Kinome Data and Virtual Screening of Identified Targets with Pan-Cancer Application

Cancer is a heterogeneous disease characterized by the uncontrolled growth and dissemination of abnormal cells. Unfortunately, it is also challenging to treat due to the heterogeneous nature of the tumors and the lack of known drugs that are effective against all tumor subpopulations. The human protein kinases represent an essential and diverse family of enzymes and are often dysregulated in cancer. The human kinome comprises only around 2% of all coding genes of the human genome but phosphorylates about 30% of cellular proteins critical for regulating various biological processes such as cell proliferation, cell cycle progression, apoptosis, motility, growth, differentiation, among others. Dysregulation of kinase activity, such as altered expression and/or amplification, aberrant phosphorylation, mutation, chromosomal translocation, and epigenetic regulation, are frequently oncogenic and can be critical for the survival and spread of cancer cells. In this dissertation, the human kinome’s genomic, epigenomic, and proteomic patterns were analyzed using the TCGA, CCLE, and GTeX databases. Unsupervised clustering based on kinome multi-omics data revealed the grouping of cancers based on their organ and tissue type. We observed significant differences in the overall kinase methylation levels (hyper- and hypomethylation) between the tumor and adjacent normal samples from the same tissue. Methylation expression quantitative trait loci (meQTL) analysis using kinase gene expression with the corresponding methylated probes revealed a highly significant and mostly negative association (~92%) within 1.5 kb from the transcription start site (TSS). Copy number alterations and mutation analysis were performed, focusing on dark kinases. We further leveraged results from multi-omics data to identify potential kinase markers of prognostic and diagnostic importance. Several understudied (dark) kinases (PKMYT1, PNCK, BRSK2, ERN2, STK31, STK32A, and MAPK4) were identified with a significant role in patient survival. Using commercially available molecules, virtual screening was performed for two dark kinase targets (PKMYT1 and PNCK). This study aims at kinase-centered analysis using multiple layers of ‘omics’ data and computational approaches to prioritize candidate kinases that could serve as potential targets from a drug development perspective