Deconvolution of DNA methylation signatures identifies common differentially methylated gene regions on 1p36 across breast cancer subtypes

ABSTRACTBreast cancer is a complex disease and studying DNA methylation (DNAm) in tumors is complicated by disease heterogeneity. We compared DNAm in breast tumors with normal-adjacent breast samples from The Cancer Genome Atlas (TCGA). We constructed models stratified by tumor stage and PAM50 molecular subtype and performed cell-type reference-free deconvolution on each model. We identified nineteen differentially methylated gene regions (DMGRs) in early stage tumors across eleven genes (AGRN, C1orf170, FAM41C, FLJ39609, HES4, ISG15, KLHL17, NOC2L, PLEKHN1, SAMD11, WASH5P). These regions were consistently differentially methylated in every subtype and all implicated genes are localized on chromosome 1p36.3. We also validated seventeen DMGRs in an independent data set. Identification and validation of shared DNAm alterations across tumor subtypes in early stage tumors advances our understanding of common biology underlying breast carcinogenesis and may contribute to biomarker development. We also provide evidence on the importance and potential function of 1p36 in cancer.

Detection of differentially methylated gene promoters in failing and nonfailing human left ventricle myocardium using computation analysis

Human dilated cardiomyopathy (DCM) is characterized by congestive heart failure and altered myocardial gene expression. Epigenetic changes, including DNA methylation, are implicated in the development of DCM but have not been studied extensively. Clinical human DCM and nonfailing control left ventricle samples were individually analyzed for DNA methylation and expressional changes. Expression microarrays were used to identify 393 overexpressed and 349 underexpressed genes in DCM (GEO accession number: GSE43435 ). Gene promoter microarrays were utilized for DNA methylation analysis, and the resulting data were analyzed by two different computational methods. In the first method, we utilized subtractive analysis of DNA methylation peak data to identify 158 gene promoters exhibiting DNA methylation changes that correlated with expression changes. In the second method, a two-stage approach combined a particle swarm optimization feature selection algorithm and a discriminant analysis via mixed integer programming classifier to identify differentially methylated gene promoters. This analysis identified 51 hypermethylated promoters and six hypomethylated promoters in DCM with 100% cross-validation accuracy in the group assignment. Generation of a composite list of genes identified by subtractive analysis and two-stage computation analysis revealed four genes that exhibited differential DNA methylation by both methods in addition to altered gene expression. Computationally identified genes ( AURKB, BTNL9, CLDN5, and TK1) define a central set of differentially methylated gene promoters that are important in classifying DCM. These genes have no previously reported role in DCM. This study documents that rigorous computational analysis applied to microarray analysis of healthy and diseased human heart samples helps to define clinically relevant DNA methylation and expressional changes in DCM.