Cancer is a complex disease characterized by molecular heterogeneity and the involvement of several cellular mechanisms throughout its evolution and pathogenesis. Despite the great efforts made to untangle these mechanisms, cancer pathophysiology remains far from clear. So far, panels of biomarkers have been reported from high-throughput data generated through different platforms. These biomarkers are primarily focused on one type of coding molecules such as transcripts or proteins, mainly due to the apparent heterogeneity of output data resulting from the use of various techniques specific to the molecular type. Hence, there is a major need to understand how these molecules interact and complement each other to be able to explain the deregulated processes involved. The breadth of large-scale data availability as well as the lack of in-depth analysis of publicly available data has raised concerns and enabled opportunities for new strategies to analyze "Big data" more comprehensively. Here, a new protocol to perform integrative analysis based on a systems biology approach is described. The foundation of the approach relies on groups of datasets from published studies compared within the original described groups and organized in a designated format to allow the integration and cross-comparison among different studies and different platforms. This approach follows an unbiased hypothesis-free methodology that will facilitate the identification of commonalities among different data-set sources, and ultimately map and characterize specific molecular pathways using significantly deregulated molecules. This in turn will generate novel insights about the mechanisms deregulated in complex diseases such as cancer.
Keywords: Cancer; High-throughput data; Integrative analysis; Molecular biomarkers; OMICS; Systems biology.