Integration of datasets for individual prediction of DNA methylation-based biomarkers.

BACKGROUND: Epigenetic scores (EpiScores) can provide biomarkers of lifestyle and disease risk. Projecting new datasets onto a reference panel is challenging due to separation of technical and biological variation with array data. Normalisation can standardise data distributions but may also remove population-level biological variation. RESULTS: We compare two birth cohorts (Lothian Birth Cohorts of 1921 and 1936 - nLBC1921 = 387 and nLBC1936 = 498) with blood-based DNA methylation assessed at the same chronological age (79 years) and processed in the same lab but in different years and experimental batches. We examine the effect of 16 normalisation methods on a novel BMI EpiScore (trained in an external cohort, n = 18,413), and Horvath's pan-tissue DNA methylation age, when the cohorts are normalised separately and together. The BMI EpiScore explains a maximum variance of R2=24.5% in BMI in LBC1936 (SWAN normalisation). Although there are cross-cohort R2 differences, the normalisation method makes a minimal difference to within-cohort estimates. Conversely, a range of absolute differences are seen for individual-level EpiScore estimates for BMI and age when cohorts are normalised separately versus together. While within-array methods result in identical EpiScores whether a cohort is normalised on its own or together with the second dataset, a range of differences is observed for between-array methods. CONCLUSIONS: Normalisation methods returning similar EpiScores, whether cohorts are analysed separately or together, will minimise technical variation when projecting new data onto a reference panel. These methods are important for cases where raw data is unavailable and joint normalisation of cohorts is computationally expensive.

Development and validation of DNA methylation scores in two European cohorts augment 10-year risk prediction of type 2 diabetes.

Type 2 diabetes mellitus (T2D) presents a major health and economic burden that could be alleviated with improved early prediction and intervention. While standard risk factors have shown good predictive performance, we show that the use of blood-based DNA methylation information leads to a significant improvement in the prediction of 10-year T2D incidence risk. Previous studies have been largely constrained by linear assumptions, the use of cytosine-guanine pairs one-at-a-time and binary outcomes. We present a flexible approach (via an R package, MethylPipeR) based on a range of linear and tree-ensemble models that incorporate time-to-event data for prediction. Using the Generation Scotland cohort (training set ncases = 374, ncontrols = 9,461; test set ncases = 252, ncontrols = 4,526) our best-performing model (area under the receiver operating characteristic curve (AUC) = 0.872, area under the precision-recall curve (PRAUC) = 0.302) showed notable improvement in 10-year onset prediction beyond standard risk factors (AUC = 0.839, precision-recall AUC = 0.227). Replication was observed in the German-based KORA study (n = 1,451, ncases = 142, P = 1.6 × 10-5).

Data-driven design of targeted gene panels for estimating immunotherapy biomarkers.

Tumour mutation burden and other exome-wide biomarkers are used to determine which patients will benefit from immunotherapy. However, the cost of whole exome sequencing limits the widespread use of such biomarkers. Here, we introduce a data-driven framework for the design of targeted gene panels for estimating a broad class of biomarkers including tumour mutation burden and tumour indel burden. Our first goal is to develop a generative model for the profile of mutation across the exome, which allows for gene- and variant type-dependent mutation rates. Based on this model, we then propose a procedure for constructing biomarker estimators. Our approach allows the practitioner to select a targeted gene panel of prespecified size and construct an estimator that only depends on the selected genes. Alternatively, our method may be applied to make predictions based on an existing gene panel, or to augment a gene panel to a given size. We demonstrate the excellent performance of our proposal using data from three non small-cell lung cancer studies, as well as data from six other cancer types.