Detection of genetic aberrations in cancer driving signaling pathways based on joint analysis of heterogeneous genomics data

Cancer results from genetic aberrations that affect multiple intracellular processes, combined with evolution and clonal selection that give cancer a significant advantage over heathy cells. Identification of genes and processes that function improperly in cancer cells is a significant challenge, necessary for the proper understanding of cancerogenesis and for the development of successful treatment scenarios. This paper shows the advantages of utilizing data provided by various methods to complement the knowledge about alterations in regulatory processes associated with cancer. Using four different experiment types that focus on mutations and indels, gene expression, copy number variation and methylation, used to study the genome of over 2000 patients treated for breast, thyroid and prostate cancer, we test some of the assumptions used in cancer research associated with coverage and mutual exclusivity of alterations. We show that individual methods do not allow to observe alterations in all cancer related processes and that the exclusivity assumption is valid only for individual alteration types. We additionally show the relationship between the violation of those assumptions and clinical data, at the level of individual patients, providing a comprehensive description of the analysis strategy used and its possible impact on the interpretation of data.