Decoding pan-cancer complexity. Multiomic insights from the lung and breast

Cancer presently represents a significant global health challenge, illustrated by the high incidence and mortality rates associated with lung and breast cancer. Technological advances and concerted large-scale initiatives have yielded vast amounts of cancer data to be used for research purposes. The objective of this thesis is to capitalize on the current state of the bioinformatics field to develop and employ tools that extract biological insights from multiomic data across a range of cancer types. Utilizing publicly available data sets from our research group and from other researchers worldwide, two of the included studies were centered on methodological development. Firstly, we successfully applied a gene-expression-based prognostic tool originally developed for lung adenocarcinoma, the most common histological subtype of lung cancer, to multiple cancer types, demonstrating the broad applicability of such classifiers (Paper I). Secondly, we refined a method for adjusting DNA methylation data based on the mixture of malignant and non-malignant cells in tumor samples, enhancing biological interpretability of such methylation data sets (Paper III). The remaining studies investigated the biological heterogeneity within lung and breast tumors. Specifically, we stratified breast cancer patients from southern Sweden according to whether they had had genes associated with increased breast cancer risk screened for variants and compared the two resulting groups, obtaining a real-world read-out of the current screening guidelines and patient selection criteria (Paper II). Lastly, we subdivided lung adenocarcinoma into four distinct subgroups based on adjusted DNA methylation data and characterized the resulting sample clusters, both showing congruence to previously proposed mRNA and protein subtypes and providing novel insights into this malignancy (Paper IV). Taken together, the findings presented in this thesis have contributed to our collective understanding of the complex cancer biology landscape.