Sampling uncertainty versus method uncertainty: A general framework with applications to omics biomarker selection

Uncertainty is a crucial issue in statistics which can be considered from different points of view. One type of uncertainty, typically referred to as sampling uncertainty, arises through the variability of results obtained when the same analysis strategy is applied to different samples. Another type of uncertainty arises through the variability of results obtained when using the same sample but different analysis strategies addressing the same research question. We denote this latter type of uncertainty as method uncertainty. It results from all the choices to be made for an analysis, for example, decisions related to data preparation, method choice, or model selection. In medical sciences, a large part of omics research is focused on the identification of molecular biomarkers, which can either be performed through ranking or by selection from among a large number of candidates. In this paper, we introduce a general resampling-based framework to quantify and compare sampling and method uncertainty. For illustration, we apply this framework to different scenarios related to the selection and ranking of omics biomarkers in the context of acute myeloid leukemia: variable selection in multivariable regression using different types of omics markers, the ranking of biomarkers according to their predictive performance, and the identification of differentially expressed genes from RNA-seq data. For all three scenarios, our findings suggest highly unstable results when the same analysis strategy is applied to two independent samples, indicating high sampling uncertainty and a comparatively smaller, but non-negligible method uncertainty, which strongly depends on the methods being compared.     

piRNAs and epigenetic conversion in Drosophila

Transposable element (TE) activity is repressed in the Drosophila germline by Piwi-Interacting RNAs (piRNAs), a class of small non-coding RNAs. These piRNAs are produced by discrete genomic loci containing TE fragments. In a recent publication, we tested for the existence of a strict epigenetic induction of piRNA production capacity by a locus in the D. melanogaster genome. We used 2 lines carrying a transgenic 7-copy tandem cluster (P-lacZ-white) at the same genomic site. This cluster generates in both lines a local heterochromatic sector. One line (T-1) produces high levels of ovarian piRNAs homologous to the P-lacZ-white transgenes and shows a strong capacity to repress homologous sequences in trans, whereas the other line (BX2) is devoid of both of these capacities. The properties of these 2 lines are perfectly stable over generations. We have shown that the maternal transmission of a cytoplasm carrying piRNAs from the first line can confer to the inert transgenic locus of the second, a totally de novo capacity to produce high levels of piRNAs as well as the ability to induce homology-dependent silencing in trans. These new properties are stably inherited over generations (n > 50). Furthermore, the converted locus has itself become able to convert an inert transgenic locus via cytoplasmic maternal inheritance. This results in a stable epigenetic conversion process, which can be performed recurrently—a phenomenon termed paramutation and discovered in Maize 60 y ago. Paramutation in Drosophila corresponds to the first stable paramutation in animals and provides a model system to investigate the epigenetically induced emergence of a piRNA-producing locus, a crucial step in epigenome shaping. In this Extra View, we discuss some additional functional aspects and the possible molecular mechanism of this piRNA-linked paramutation.     

Polycystic ovary syndrome is transmitted via a transgenerational epigenetic process

1. Download : Download high-res image (145KB)
2. Download : Download full-size image

     

Structural characterization of the α-N-acetylglucosaminidase,a key enzyme in the pathogenesis of Sanfilippo syndrome B

Mucopolysaccharidosis III B (MPS III-B) is a rare lysosomal storage disorder caused by deficiencies in Alpha-N-acetylglucosaminidase (NAGLU) for which there is currently no cure, and present treatment is largely supportive. Understanding the structure of NAGLU may allow for identification of novel therapeutic targets for MPS III-B. Here we describe the first crystal structure of human NAGLU, determined to a resolution of 2.3?Å. The crystal structure reveals a novel homotrimeric configuration, maintained primarily by hydrophobic and electrostatic interactions via domain II of three contiguous domains from the N- to C-terminus. The active site cleft is located between domains II and III. Catalytic glutamate residues, E316 and E446, are located at the top of the (α/β)₈ barrel structure in domain II. We utilized the three-dimensional structure of NAGLU to map several MPS III-B mutations, and hypothesize their functional consequences. Revealing atomic level structural information about this critical lysosomal enzyme paves the way for the design of novel therapeutics to target the underlying causes of MPS III-B.     

Genomic and RT-qPCR analysis of trimethoprim-sulfamethoxazole and meropenem resistance in Burkholderia pseudomallei clinical isolates

BackgroundMelioidosis is an endemic disease in southeast Asia and northern Australia caused by the saprophytic bacteria Burkholderia pseudomallei, with a high mortality rate. The clinical presentation is multifaceted, with symptoms ranging from acute septicemia to multiple chronic abscesses. Here, we report a chronic case of melioidosis in a patient who lived in Malaysia in the 70s and was suspected of contracting tuberculosis. Approximately 40 years later, in 2014, he was diagnosed with pauci-symptomatic melioidosis during a routine examination. Four strains were isolated from a single sample. They showed divergent morphotypes and divergent antibiotic susceptibility, with some strains showing resistance to trimethoprim-sulfamethoxazole and fluoroquinolones. In 2016, clinical samples were still positive for B. pseudomallei, and only one type of strain, showing atypical resistance to meropenem, was isolated.Principal findingsWe performed whole genome sequencing and RT-qPCR analysis on the strains isolated during this study to gain further insights into their differences. We thus identified two types of resistance mechanisms in these clinical strains. The first one was an adaptive and transient mechanism that disappeared during the course of laboratory sub-cultures; the second was a mutation in the efflux pump regulator amrR, associated with the overexpression of the related transporter.ConclusionThe development of such mechanisms may have a clinical impact on antibiotic treatment. Indeed, their transient nature could lead to an undiagnosed resistance. Efflux overexpression due to mutation leads to an important multiple resistance, reducing the effectiveness of antibiotics during treatment.     

Determining the Cellular Structure of Two Cereal Food Foams by X-ray Micro-Tomography

The cellular structure of two products, an extruded breakfast cereal and a short dough biscuit, was characterized by two different X-ray micro computed tomographic systems. Acquisitions were made by a compact desktop system Skyscan 1174 (Bruker μCT, Belgium) and at the European Synchrotron Radiation Facility (ESRF, beamline ID19, France) at different resolutions (voxel size of 6.5 μm, 7.5 μm, 16.2 μm and 25.8 μm). 3D images were processed for the density, the connectivity index and the granulometry of cells and cell walls. These experiments underlined the importance of the resolution for determination of quantitative measurements such as densities and thicknesses. The median width calculated for the cell walls distribution in the biscuit dropped from 141 to 50 μm when the voxel size changed from 25.8 to 7.5 μm. Images well showed that even though the food products had close values of porosity 0.6 and 0.7 for biscuit and extruded breakfast cereal respectively, their cellular structures were very different. The biscuit had small cells (median value of the distribution varied from 125 to 152 μm, according to resolution) and larger cell walls (50–141 μm) than the extrudate (32–109 μm) which, on the contrary, exhibited very large cells (307–400 μm). Beyond methodological issues, these differences could be clearly attributed to the differences of compositions and processes.