John A. Dawson

EBSeq: An empirical Bayes hierarchical model for inference in RNA-seq experiments

MOTIVATION Messenger RNA expression is important in normal development and differentiation, as well as in manifestation of disease. RNA-seq experiments allow for the identification of differentially expressed (DE) genes and their corresponding isoforms on a genome-wide scale. However, statistical methods are required to ensure that accurate identifications are made. A number of methods exist for identifying DE genes, but far fewer are available for identifying DE isoforms. When isoform DE is of interest, investigators often apply gene-level (count-based) methods directly to estimates of isoform counts. Doing so is not recommended. In short, estimating isoform expression is relatively straightforward for some groups of isoforms, but more challenging for others. This results in estimation uncertainty that varies across isoform groups. Count-based methods were not designed to accommodate this varying uncertainty, and consequently, application of them for isoform inference results in reduced power for some classes of isoforms and increased false discoveries for others. RESULTS Taking advantage of the merits of empirical Bayesian methods, we have developed EBSeq for identifying DE isoforms in an RNA-seq experiment comparing two or more biological conditions. Results demonstrate substantially improved power and performance of EBSeq for identifying DE isoforms. EBSeq also proves to be a robust approach for identifying DE genes. AVAILABILITY AND IMPLEMENTATION An R package containing examples and sample datasets is available at http://www.biostat.wisc.edu/kendzior/EBSEQ/. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

The SIRT1 activator SRT1720 extends lifespan and improves health of mice fed a standard diet

The prevention or delay of the onset of age-related diseases prolongs survival and improves quality of life while reducing the burden on the health care system. Activation of sirtuin 1 (SIRT1), an NAD(+)-dependent deacetylase, improves metabolism and confers protection against physiological and cognitive disturbances in old age. SRT1720 is a specific SIRT1 activator that has health and lifespan benefits in adult mice fed a high-fat diet. We found extension in lifespan, delayed onset of age-related metabolic diseases, and improved general health in mice fed a standard diet after SRT1720 supplementation. Inhibition of proinflammatory gene expression in both liver and muscle of SRT1720-treated animals was noted. SRT1720 lowered the phosphorylation of NF-κB pathway regulators in vitro only when SIRT1 was functionally present. Combined with our previous work, the current study further supports the beneficial effects of SRT1720 on health across the lifespan in mice.

The effectiveness of breakfast recommendations on weight loss: a randomized controlled trial

BACKGROUND Breakfast is associated with lower body weight in observational studies. Public health authorities commonly recommend breakfast consumption to reduce obesity, but the effectiveness of adopting these recommendations for reducing body weight is unknown. OBJECTIVE We tested the relative effectiveness of a recommendation to eat or skip breakfast on weight loss in adults trying to lose weight in a free-living setting. DESIGN We conducted a multisite, 16-wk, 3-parallel-arm randomized controlled trial in otherwise healthy overweight and obese adults [body mass index (in kg/m²) between 25 and 40] aged 20-65 y. Our primary outcome was weight change. We compared weight change in a control group with weight loss in experimental groups told to eat breakfast or to skip breakfast [no breakfast (NB)]. Randomization was stratified by prerandomization breakfast eating habits. A total of 309 participants were randomly assigned. RESULTS A total of 283 of the 309 participants who were randomly assigned completed the intervention. Treatment assignment did not have a significant effect on weight loss, and there was no interaction between initial breakfast eating status and treatment. Among skippers, mean (±SD) baseline weight-, age-, sex-, site-, and race-adjusted weight changes were -0.71 ± 1.16, -0.76 ± 1.26, and -0.61 ± 1.18 kg for the control, breakfast, and NB groups, respectively. Among breakfast consumers, mean (±SD) baseline weight-, age-, sex-, site-, and race-adjusted weight changes were -0.53 ± 1.16, -0.59 ± 1.06, and -0.71 ± 1.17 kg for the control, breakfast, and NB groups, respectively. Self-reported compliance with the recommendation was 93.6% for the breakfast group and 92.4% for the NB group. CONCLUSIONS A recommendation to eat or skip breakfast for weight loss was effective at changing self-reported breakfast eating habits, but contrary to widely espoused views this had no discernable effect on weight loss in free-living adults who were attempting to lose weight.

Weighing the Evidence of Common Beliefs in Obesity Research

Obesity is a topic on which many views are strongly held in the absence of scientific evidence to support those views, and some views are strongly held despite evidence to contradict those views. We refer to the former as “presumptions” and the latter as “myths.” Here, we present nine myths and 10 presumptions surrounding the effects of rapid weight loss; setting realistic goals in weight loss therapy; stage of change or readiness to lose weight; physical education classes; breastfeeding; daily self-weighing; genetic contribution to obesity; the “Freshman 15”; food deserts; regularly eating (versus skipping) breakfast; eating close to bedtime; eating more fruits and vegetables; weight cycling (i.e., yo-yo dieting); snacking; built environment; reducing screen time in childhood obesity; portion size; participation in family mealtime; and drinking water as a means of weight loss. For each of these, we describe the belief and present evidence that the belief is widely held or stated, reasons to support the conjecture that the belief might be true, evidence to directly support or refute the belief, and findings from randomized controlled trials, if available. We conclude with a discussion of the implications of these determinations, conjecture on why so many myths and presumptions exist, and suggestions for limiting the spread of these and other unsubstantiated beliefs about the obesity domain.

Predicting adult weight change in the real world: a systematic review and meta-analysis accounting for compensatory changes in energy intake or expenditure

Background:Public health and clinical interventions for obesity in free-living adults may be diminished by individual compensation for the intervention. Approaches to predict weight outcomes do not account for all mechanisms of compensation, so they are not well suited to predict outcomes in free-living adults. Our objective was to quantify the range of compensation in energy intake or expenditure observed in human randomized controlled trials (RCTs).Methods:We searched multiple databases (PubMed, CINAHL, SCOPUS, Cochrane, ProQuest, PsycInfo) up to 1 August 2012 for RCTs evaluating the effect of dietary and/or physical activity interventions on body weight/composition. Inclusion criteria: subjects per treatment arm ⩾5; ⩾1 week intervention; a reported outcome of body weight/body composition; the intervention was either a prescribed amount of over- or underfeeding and/or supervised or monitored physical activity was prescribed; ⩾80% compliance; and an objective method was used to verify compliance with the intervention (for example, observation and electronic monitoring). Data were independently extracted and analyzed by multiple reviewers with consensus reached by discussion. We compared observed weight change with predicted weight change using two models that predict weight change accounting only for metabolic compensation.Findings:Twenty-eight studies met inclusion criteria. Overfeeding studies indicate 96% less weight gain than expected if no compensation occurred. Dietary restriction and exercise studies may result in up to 12–44% and 55–64% less weight loss than expected, respectively, under an assumption of no behavioral compensation.Interpretation:Compensation is substantial even in high-compliance conditions, resulting in far less weight change than would be expected. The simple algorithm we report allows for more realistic predictions of intervention effects in free-living populations by accounting for the significant compensation that occurs.

Cholecystokinin Is Up-Regulated in Obese Mouse Islets and Expands β-Cell Mass by Increasing β-Cell Survival

An absolute or functional deficit in beta-cell mass is a key factor in the pathogenesis of diabetes. We model obesity-driven beta-cell mass expansion by studying the diabetes-resistant C57BL/6-Leptin(ob/ob) mouse. We previously reported that cholecystokinin (Cck) was the most up-regulated gene in obese pancreatic islets. We now show that islet cholecystokinin (CCK) is up-regulated 500-fold by obesity and expressed in both alpha- and beta-cells. We bred a null Cck allele into the C57BL/6-Leptin(ob/ob) background and investigated beta-cell mass and metabolic parameters of Cck-deficient obese mice. Loss of CCK resulted in decreased islet size and reduced beta-cell mass through increased beta-cell death. CCK deficiency and decreased beta-cell mass exacerbated fasting hyperglycemia and reduced hyperinsulinemia. We further investigated whether CCK can directly affect beta-cell death in cell culture and isolated islets. CCK was able to directly reduce cytokine- and endoplasmic reticulum stress-induced cell death. In summary, CCK is up-regulated by islet cells during obesity and functions as a paracrine or autocrine factor to increase beta-cell survival and expand beta-cell mass to compensate for obesity-induced insulin resistance.

Reference Values for Body Composition and Anthropometric Measurements in Athletes

Background Despite the importance of body composition in athletes, reference sex- and sport-specific body composition data are lacking. We aim to develop reference values for body composition and anthropometric measurements in athletes. Methods Body weight and height were measured in 898 athletes (264 female, 634 male), anthropometric variables were assessed in 798 athletes (240 female and 558 male), and in 481 athletes (142 female and 339 male) with dual-energy X-ray absorptiometry (DXA). A total of 21 different sports were represented. Reference percentiles (5th, 25th, 50th, 75th, and 95th) were calculated for each measured value, stratified by sex and sport. Because sample sizes within a sport were often very low for some outcomes, the percentiles were estimated using a parametric, empirical Bayesian framework that allowed sharing information across sports. Results We derived sex- and sport-specific reference percentiles for the following DXA outcomes: total (whole body scan) and regional (subtotal, trunk, and appendicular) bone mineral content, bone mineral density, absolute and percentage fat mass, fat-free mass, and lean soft tissue. Additionally, we derived reference percentiles for height-normalized indexes by dividing fat mass, fat-free mass, and appendicular lean soft tissue by height squared. We also derived sex- and sport-specific reference percentiles for the following anthropometry outcomes: weight, height, body mass index, sum of skinfold thicknesses (7 skinfolds, appendicular skinfolds, trunk skinfolds, arm skinfolds, and leg skinfolds), circumferences (hip, arm, midthigh, calf, and abdominal circumferences), and muscle circumferences (arm, thigh, and calf muscle circumferences). Conclusions These reference percentiles will be a helpful tool for sports professionals, in both clinical and field settings, for body composition assessment in athletes.

Metabolic Changes in Skin Caused by Scd1 Deficiency: A Focus on Retinol Metabolism

We previously reported that mice with skin-specific deletion of stearoyl-CoA desaturase-1 (Scd1) recapitulated the skin phenotype and hypermetabolism observed in mice with a whole-body deletion of Scd1. In this study, we first performed a diet-induced obesity experiment at thermoneutral temperature (33°C) and found that skin-specific Scd1 knockout (SKO) mice still remain resistant to obesity. To elucidate the metabolic changes in the skin that contribute to the obesity resistance and skin phenotype, we performed microarray analysis of skin gene expression in male SKO and control mice fed a standard rodent diet. We identified an extraordinary number of differentially expressed genes that support the previously documented histological observations of sebaceous gland hypoplasia, inflammation and epidermal hyperplasia in SKO mice. Additionally, transcript levels were reduced in skin of SKO mice for genes involved in fatty acid synthesis, elongation and desaturation, which may be attributed to decreased abundance of key transcription factors including SREBP1c, ChREBP and LXRα. Conversely, genes involved in cholesterol synthesis were increased, suggesting an imbalance between skin fatty acid and cholesterol synthesis. Unexpectedly, we observed a robust elevation in skin retinol, retinoic acid and retinoic acid-induced genes in SKO mice. Furthermore, SEB-1 sebocytes treated with retinol and SCD inhibitor also display an elevation in retinoic acid-induced genes. These results highlight the importance of monounsaturated fatty acid synthesis for maintaining retinol homeostasis and point to disturbed retinol metabolism as a novel contributor to the Scd1 deficiency-induced skin phenotype.

An Empirical Bayesian Approach for Identifying Differential Coexpression in High-Throughput Experiments

A common goal of microarray and related high-throughput genomic experiments is to identify genes that vary across biological condition. Most often this is accomplished by identifying genes with changes in mean expression level, so called differentially expressed (DE) genes, and a number of effective methods for identifying DE genes have been developed. Although useful, these approaches do not accommodate other types of differential regulation. An important example concerns differential coexpression (DC). Investigations of this class of genes are hampered by the large cardinality of the space to be interrogated as well as by influential outliers. As a result, existing DC approaches are often underpowered, exceedingly prone to false discoveries, and/or computationally intractable for even a moderately large number of pairs. To address this, an empirical Bayesian approach for identifying DC gene pairs is developed. The approach provides a false discovery rate controlled list of significant DC gene pairs without sacrificing power. It is applicable within a single study as well as across multiple studies. Computations are greatly facilitated by a modification to the expectation-maximization algorithm and a procedural heuristic. Simulations suggest that the proposed approach outperforms existing methods in far less computational time; and case study results suggest that the approach will likely prove to be a useful complement to current DE methods in high-throughput genomic studies.

R/EBcoexpress

UNLABELLED R/EBcoexpress implements the approach of Dawson and Kendziorski using R, a freely available, open source statistical programming language. The approach identifies differential co-expression (DC) by examining the correlations among gene pairs using an empirical Bayesian approach, producing a false discovery rate controlled list of DC pairs. This interrogation of DC gene pairs complements but is distinct from differential expression analyses, under the general goal of understanding differential regulation across biological conditions. AVAILABILITY AND IMPLEMENTATION R/EBcoexpress is freely available and hosted on Bioconductor; a source file and vignette may be found at http://www.bioconductor.org/packages/release/bioc/html/EBcoexpress.html