Machine Learning Identifies Stemness Features Associated with Oncogenic Dedifferentiation.

Cancer progression involves the gradual loss of a differentiated phenotype and acquisition of progenitor and stem-cell-like features. Here, we provide novel stemness indices for assessing the degree of oncogenic dedifferentiation. We used an innovative one-class logistic regression (OCLR) machine-learning algorithm to extract transcriptomic and epigenetic feature sets derived from non-transformed pluripotent stem cells and their differentiated progeny. Using OCLR, we were able to identify previously undiscovered biological mechanisms associated with the dedifferentiated oncogenic state. Analyses of the tumor microenvironment revealed unanticipated correlation of cancer stemness with immune checkpoint expression and infiltrating immune cells. We found that the dedifferentiated oncogenic phenotype was generally most prominent in metastatic tumors. Application of our stemness indices to single-cell data revealed patterns of intra-tumor molecular heterogeneity. Finally, the indices allowed for the identification of novel targets and possible targeted therapies aimed at tumor differentiation.

The DNA methylation drift of the atherosclerotic aorta increases with lesion progression.

BACKGROUND: Atherosclerosis severity-independent alterations in DNA methylation, a reversible and highly regulated DNA modification, have been detected in aortic atheromas, thus supporting the hypothesis that epigenetic mechanisms participate in the pathogenesis of atherosclerosis. One yet unaddressed issue is whether the progression of atherosclerosis is associated with an increase in DNA methylation drift in the vascular tissue. The purpose of the study was to identify CpG methylation profiles that vary with the progression of atherosclerosis in the human aorta. METHODS: We interrogated a set of donor-matched atherosclerotic and normal aortic samples ranging from histological grade III to VII, with a high-density (>450,000 CpG sites) DNA methylation microarray. RESULTS: We detected a correlation between histological grade and intra-pair differential methylation for 1,985 autosomal CpGs, the vast majority of which drifted towards hypermethylation with lesion progression. The identified CpG loci map to genes that are regulated by known critical transcription factors involved in atherosclerosis and participate in inflammatory and immune responses. Functional relevance was corroborated by crossing the DNA methylation profiles with expression data obtained in the same human aorta sample set, by a transcriptome-wide analysis of murine atherosclerotic aortas and from available public databases. CONCLUSIONS: Our work identifies for the first time atherosclerosis progression-specific DNA methylation profiles in the vascular tissue. These findings provide potential novel markers of lesion severity and targets to counteract the progression of the atheroma.