Search

A new version of nanorate DNA sequencing, with an error rate lower than five errors per billion base pairs and compatible with whole-exome and targeted capture, enables epidemiological-scale studies of somatic mutation and selection and the generation of high-resolution selection maps across coding and non-coding sites for many genes.

CellDMC finds cell type–specific differential methylation in mixtures of cells, including epithelial tissues, and scenarios in which methylation changes in opposite directions in different cell types.

DNA methylation (DNAm) clocks can track mitotic age, but their potential use for cancer risk prediction remains less explored. Here, the authors develop a DNAm counter of total mitotic age (stemTOC) that shows an increase of mitotic age in normal tissues and precancerous lesions.

Altered epigenetics is a feature of cancer but whether these changes occur early in tumour development is unclear. Here, the authors analyse methylation events in breast cancer and adjacent normal pairs, and show that methylation changes in the normal tissue are also found in the tumour, suggesting that some of these events occur early in cancer.

Robust quantification of the differentiation potential of single cells is a task of great importance. Here the authors integrate single-cell RNA-Seq profiles with a cellular interaction network to compute the signaling entropy, and show that it can identify normal and cancer stem-cell phenotypes.

Currently, most miRNA profiling study of complex tissues are based on bulk sample sequencing, among which the cellular heterogeneity may become an important potential confounding factor. Here, the authors developed a deconvolution method to systematically decode cellular components of complex tissues from miRNA profiling data.

Tong et al. construct simulations using DNA methylation data to quantify what proportion of the predictive accuracy of epigenetic clocks could be explained by stochastic methylation changes, suggesting that stochasticity contributes more toward the accuracy of chronological rather than biological age predictions.

The incidence of type 1 diabetes is increasing, potentially implicating non-genetic factors. Here the authors conduct an epigenome-wide association study in disease-discordant twins and find increased DNA methylation variability at genes associated with immune cell metabolism and the cell cycle.

In this Perspective, Teschendorff and Feinberg describe how single-cell analysis methods based on statistical mechanics can provide valuable insights into developmental phenomena, such as differentiation potency and lineage trajectories, as well as disruption of these processes in cancer.

DNA methylation data of whole blood from Han Chinese individuals are used to map mQTLs, finding that cis- and trans-mQTLs show distinct patterns of East Asians (EA) ancestry-specific colocalization with complex trait variation.

Cancer-associated mutations in isocitrate dehydrogenase are proposed to impair TET2-dependent DNA demethylation. By comparing the methylomes of IDH-mutant cancers, the authors identify the transcription factor EBF1 as a partner of TET2, suggesting a possible means for targeting TET2 to specific DNA sequences.

Single-cell technologies allow quantification of cell-types in human tissues, yet detecting if their abundance changes in aging or disease is challenging. By using cell-attribute aware clustering, this work presents a differential abundance testing algorithm with increased power.

Feinberg and Timp review cancer-associated epigenetic alterations and propose that epigenetic dysregulation is an initiating force in tumorigenesis that promotes the selection of cancer-associated phenotypes and that can cooperate with genetic alterations, indicating that the gene-centric view of cancer biology is not the whole story.

Smoking-associated DNA methylation changes in whole blood have been reported by many EWAS. Here, the authors use a cell-type deconvolution algorithm to identify cell-type specific DNA methylation signals in seven EWAS, identifying lineage-specific smoking-associated DNA methylation changes.

TGF-β signalling suppresses tumorigenesis in breast cancer cells but its effects on breast cancer initiating cells have not been reported. Using cells in culture, Brunaet al. show that TGF-β increases breast cancer initiating cell numbers in cells that have low levels of the tight junction protein claudin.

The wealth of DNA methylation data continues to grow rapidly, including from epigenome-wide association studies (EWAS). However, extracting meaningful biological and clinical information requires diverse computational approaches for data analysis. This Review discusses the range of statistical tools available, including for cell-type deconvolution, identification of important methylation data features, causation and system-level integration with other types of omic data.

Epigenetic clocks based on DNA methylation data are machine learning tools used to estimate chronological and biological age. The authors review computational and statistical challenges that must be addressed for the rigorous construction of interpretable epigenetic clocks at cell-type and single-cell resolution.

Genome-wide mapping of oestrogen receptor-α binding sites in primary breast cancer tissues shows that oestrogen receptor binding is dynamically regulated and that the expression of genes near differentially bound regulatory regions is associated with clinical outcome.

This resource presents an in silico generated DNA methylation atlas that can be used for cell-type deconvolution of human tissues.

Transcription factors control cell identity and function in health, disease and aging. Here the authors identify age-associated changes of transcriptional regulatory networks in single cells, revealing cell-type and tissue-type-independent patterns in key pathways, including circadian rhythm, antigen processing, collagen processing and inflammation.

This Comment describes some of the common pitfalls encountered in deriving and validating predictive statistical models from high-dimensional data. It offers a fresh perspective on some key statistical issues, providing some guidelines to avoid pitfalls, and to help unfamiliar readers better assess the reliability and significance of their results.

This Perspective provides an overview of major challenges and associated recommendations in computational deconvolution for genomics data.

Advances in next generation sequencing have made it possible to precisely characterize the coding mutations that occur during the development and progression of individual cancers. Here, this technique is used to sequence the genomes and transcriptomes of an oestrogen-receptor-α-positive metastatic lobular breast cancer; significant evolution is found to occur with disease progression.

A miRNA multi-omics study reveals not only mechanisms for miRNA dysregulation, but also the tumor heterogeneity and immunological diversity within low grade glioma, and presents better prognostic value than traditional molecular markers.

In this Review, the authors give an overview of age-related changes to the methylation landscape and the statistical tools used to quantify them. They also discuss the evidence for longevity strategies that aim to counter these changes, and the candidate mechanistic causes of epigenetic ageing.

The development of cancer involves several epigenomic alterations, and the presence of certain alterations before the development of cancer is associated with cancer risk. In this Review, the authors describe the potential of epigenomics-based assays to predict an individual's risk of cancer, including discussions of technical, practical and societal issues regarding the implementation of such assays.

Chen et al build on a marker-free entropy-based algorithm to assign specific potency states to individual cells using single-cell RNA sequencing data. They identify a putative bipotent stem-cell like state, which associates with basal breast cancer risk and poor clinical outcome.