With recent advances in sequencing technology, it is now feasible to measure DNA methylation at tens of millions of sites across the entire genome. In most applications, biologists are interested in detecting differentially methylated regions, composed of multiple sites with differing methylation levels among populations. However, current computational approaches for detecting such regions do not provide accurate statistical inference. A major challenge in reporting uncertainty is that a genome-wide scan is involved in detecting these regions, which needs to be accounted for. A further challenge is that sample sizes are limited due to the costs associated with the technology. We have developed a new approach that overcomes these challenges and assesses uncertainty for differentially methylated regions in a rigorous manner. Region-level statistics are obtained by fitting a generalized least squares regression model with a nested autoregressive correlated error structure for the effect of interest on transformed methylation proportions. We develop an inferential approach, based on a pooled null distribution, that can be implemented even when as few as two samples per population are available. Here, we demonstrate the advantages of our method using both experimental data and Monte Carlo simulation. We find that the new method improves the specificity and sensitivity of lists of regions and accurately controls the false discovery rate.

Summary 10With recent advances in sequencing technology, it is now feasible to measure DNA methylation 11 at tens of millions of sites across the entire genome. In most applications, biologists are 12 interested in detecting differentially methylated regions, composed of multiple sites with 13 differing methylation levels among populations. However, current computational approaches for 14 detecting such regions do not provide accurate statistical inference. A major challenge in 15 reporting uncertainty is that a genome-wide scan is involved in detecting these regions, which 16 needs to be accounted for. A further challenge is that sample sizes are limited due to the costs 17 associated with the technology. We have developed a new approach that overcomes these 18 challenges and assesses uncertainty for differentially methylated regions in a rigorous manner. 19Region-level statistics are obtained by fitting a generalized least squares (GLS) regression model 20 with a nested autoregressive correlated error structure for the effect of interest on transformed 21 methylation proportions. We develop an inferential approach, based on a pooled null 22 distribution, that can be implemented even when as few as two samples per population are 23available. Here we demonstrate the advantages of our method using both experimental data and 24Monte Carlo simulation. We find that the new method improves the specificity and sensitivity of 25 list of regions and accurately controls the False Discovery Rate (FDR).

Wall shear stress (WSS) on anchored cells affects their responses, including cell proliferation and morphology. In this study, the effects of the directionality of pulsatile WSS on endothelial cell proliferation and morphology were investigated for cells grown in a Petri dish orbiting on a shaker platform. Time and location dependent WSS was determined by computational fluid dynamics (CFD). At low orbital speed (50 rpm), WSS was shown to be uniform (0-1 dyne/cm(2)) across the bottom of the dish, while at higher orbital speed (100 and 150 rpm), WSS remained fairly uniform near the center and fluctuated significantly (0-9 dyne/cm(2)) near the side walls of the dish. Since WSS on the bottom of the dish is two-dimensional, a new directional oscillatory shear index (DOSI) was developed to quantify the directionality of oscillating shear. DOSI approached zero for biaxial oscillatory shear of equal magnitudes near the center and approached one for uniaxial pulsatile shear near the wall, where large tangential WSS dominated a much smaller radial component. Near the center (low DOSI), more, smaller and less elongated cells grew, whereas larger cells with greater elongation were observed in the more uniaxial oscillatory shear (high DOSI) near the periphery of the dish. Further, cells aligned with the direction of the largest component of shear but were randomly oriented in low magnitude biaxial shear. Statistical analyses of the individual and interacting effects of multiple factors (DOSI, shear magnitudes and orbital speeds) showed that DOSI significantly affected all the responses, indicating that directionality is an important determinant of cellular responses.

We compare the performance of our method SVA-PLS with standard ANOVA and a relatively recent technique of surrogate variable analysis (SVA), on a wide variety of simulation settings (incorporating different effects of the hidden variable, under situations with varying signal intensities and gene groupings). In all settings, our method yields the highest sensitivity while maintaining relatively reasonable values for the specificity, false discovery rate and false non-discovery rate. Application of our method to gene expression profiling for acute megakaryoblastic leukemia shows that our method detects an additional six genes, that are missed by both the standard ANOVA method as well as SVA, but may be relevant to this disease, as can be seen from mining the existing literature.

No abstract