Improving replicability in single-cell RNA-Seq cell type discovery with Dune.

BACKGROUND: Single-cell transcriptome sequencing (scRNA-Seq) has allowed new types of investigations at unprecedented levels of resolution. Among the primary goals of scRNA-Seq is the classification of cells into distinct types. Many approaches build on existing clustering literature to develop tools specific to single-cell. However, almost all of these methods rely on heuristics or user-supplied parameters to control the number of clusters. This affects both the resolution of the clusters within the original dataset as well as their replicability across datasets. While many recommendations exist, in general, there is little assurance that any given set of parameters will represent an optimal choice in the trade-off between cluster resolution and replicability. For instance, another set of parameters may result in more clusters that are also more replicable. RESULTS: Here, we propose Dune, a new method for optimizing the trade-off between the resolution of the clusters and their replicability. Our method takes as input a set of clustering results-or partitions-on a single dataset and iteratively merges clusters within each partitions in order to maximize their concordance between partitions. As demonstrated on multiple datasets from different platforms, Dune outperforms existing techniques, that rely on hierarchical merging for reducing the number of clusters, in terms of replicability of the resultant merged clusters as well as concordance with ground truth. Dune is available as an R package on Bioconductor: https://www.bioconductor.org/packages/release/bioc/html/Dune.html . CONCLUSIONS: Cluster refinement by Dune helps improve the robustness of any clustering analysis and reduces the reliance on tuning parameters. This method provides an objective approach for borrowing information across multiple clusterings to generate replicable clusters most likely to represent common biological features across multiple datasets.

Does sweetness exposure drive 'sweet tooth'?

It is widely believed that exposure to sweetened foods and beverages stimulates the liking and desire for sweetness. Here we provide an updated review of the empirical evidence from human research examining whether exposure to sweet foods or beverages influences subsequent general liking for sweetness ('sweet tooth'), based on the conclusions of existing systematic reviews and more recent research identified from a structured search of literature. Prior reviews have concluded that the evidence for a relationship between sweet taste exposure and measures of sweet taste liking is equivocal, and more recent primary research generally does not support the view that exposure drives increased liking for sweetness, in adults or children. In intervention trials using a range of designs, acute exposure to sweetness usually has the opposite effect (reducing subsequent liking and desire for sweet taste), while sustained exposures have no significant effects or inconsistent effects. Recent longitudinal observational studies in infants and children also report no significant associations between exposures to sweet foods and beverages with measures of sweet taste preferences. Overall, while it is widely assumed that exposure to sweetness stimulates a greater liking and desire for sweetness, this is not borne out by the balance of empirical evidence. While new research may provide a more robust evidence base, there are also a number of methodological, biological and behavioural considerations that may underpin the apparent absence of a positive relationship between sweetness exposure and liking.

A Bartlett-type correction for likelihood ratio tests with application to testing equality of Gaussian graphical models.

This work defines a new correction for the likelihood ratio test for a two-sample problem within the multivariate normal context. This correction applies to decomposable graphical models, where testing equality of distributions can be decomposed into lower dimensional problems.

Gastrointestinal tolerance of D-allulose in children: an acute, randomised, double-blind, placebo-controlled, cross-over study.

D-Allulose, a low-calorie sugar, provides an attractive alternative to added sugars in food and beverage products. There is however limited data on its gastrointestinal (GI) tolerance, with only two studies in adults, and no studies in children to date. We therefore performed an acute, randomised, double-blind, placebo-controlled, cross over study designed to determine, for the first time, the GI tolerance of 2 doses of D-allulose (2.5 g per 120 ml and 4.3 g per 120 ml) in young children. The primary tolerance endpoint was the difference in the number of participants experiencing at least one stool that met a Type 6 or Type 7 description on the Bristol Stool Chart, within 24 hours after study product intake. Secondary endpoints included the assessment of stool frequency, stool consistency, and the presence of GI symptoms. Only one participant in the low dose group experienced a stool type 6 or 7, while no participants experienced a stool type 6 or 7 in the high dose group. A statistically significant difference in the change in stool frequency compared to placebo in the high dose group (p = 0.044) was found, with no significant difference between the groups for stool consistency and no participants experienced unusual stool frequency. All the encountered adverse events were non-serious, either mild or moderate, and there were no serious adverse events. All in all, D-allulose was tolerated well in children, making this ingredient a good candidate to reformulate commercially produced goods by replacing added sugars with lower caloric content.

Association between Variants of the TRPV1 Gene and Body Composition in Sub-Saharan Africans.

In humans, the transient receptor potential vanilloid 1 (TRPV1) gene is activated by exogenous (e.g., high temperatures, irritating compounds such as capsaicin) and endogenous (e.g., endocannabinoids, inflammatory factors, fatty acid metabolites, low pH) stimuli. It has been shown to be involved in several processes including nociception, thermosensation, and energy homeostasis. In this study, we investigated the association between TRPV1 gene variants, sensory perception (to capsaicin and PROP), and body composition (BMI and bioimpedance variables) in human populations. By comparing sequences deposited in worldwide databases, we identified two haplotype blocks (herein referred to as H1 and H2) that show strong stabilizing selection signals (MAF approaching 0.50, Tajima's D > +4.5) only in individuals with sub-Saharan African ancestry. We therefore studied the genetic variants of these two regions in 46 volunteers of sub-Saharan descent and 45 Italian volunteers (both sexes). Linear regression analyses showed significant associations between TRPV1 diplotypes and body composition, but not with capsaicin perception. Specifically, in African women carrying the H1-b and H2-b haplotypes, a higher percentage of fat mass and lower extracellular fluid retention was observed, whereas no significant association was found in men. Our results suggest the possible action of sex-driven balancing selection at the non-coding sequences of the TRPV1 gene, with adaptive effects on water balance and lipid deposition.

Modelling the public health benefits of fibre fortification in the Chinese population through food reformulation.

OBJECTIVES: Various studies have highlighted how consuming adequate dietary fibre (DF) foods could confer multiple potential health benefits to humans, though data suggested that the average intake of the population is below the recommendations. The aim of this study, which involved probabilistic, mathematical and statistical modelling, was to understand, for the first time, how fibre fortification in a broad array of food categories could impact the diet and health status of Chinese consumers. DESIGN: A simulation-based approach was used to examine the potential impact of fibre fortification. The China Health and Nutrition Survey dataset was used to evaluate intakes of DF together with a dietary intake mathematical model. Commercially manufactured foods and beverages eligible for fibre fortification were identified and a total of 296 food and beverages were selected for fibre fortification calculation. Foods and beverages eligible for fibre fortification and the concentration of fibre used at intervention were identified based on Chinese legislations and regulations of nutrition label claims. Populations who meet the dietary reference values of fibre fortification have their health outcomes such as weight, cardiovascular disease (CVD) and type 2 diabetes risk quantified prefibre and postfibre reformulation as per published studies. RESULTS: The simulated fibre fortification intervention model has shown that the mean DF intake increased by 13.28%, from 12.8 g/day of baseline to 14.5 g/day, leading to an increase of 48% (from 6.85% to 10.13%) and 54% (from 14.22% to 21.84%) of the adult and children population, respectively, achieving the recommended fibre guidelines. Additionally, 234 diabetes cases per day (85 340 cases per year) as well as 73 065 deaths secondary to CVD could also potentially be averted or delayed with the increase of DF intake via fibre fortification. CONCLUSIONS: This study provides a practical application implicating the potential public health benefits that could be achieved with food product reformulation.

Differential detection workflows for multi-sample single-cell RNA-seq data.

In single-cell transcriptomics, differential gene expression (DE) analyses typically focus on testing differences in the average expression of genes between cell types or conditions of interest. Single-cell transcriptomics, however, also has the promise to prioritise genes for which the expression differ in other aspects of the distribution. Here we develop a workflow for assessing differential detection (DD), which tests for differences in the average fraction of samples or cells in which a gene is detected. After benchmarking eight different DD data analysis strategies, we provide a unified workflow for jointly assessing DE and DD. Using simulations and two case studies, we show that DE and DD analysis provide complementary information, both in terms of the individual genes they report and in the functional interpretation of those genes.