Intensity-Specific Differential Leukocyte DNA Methylation in Physical (In)Activity: An Exploratory Approach.

The aim of this exploratory study was to investigate how sedentary behavior (SB) and physical activity (PA) influence DNA methylation at a global, gene-specific, and health-related pathway level. SB, light PA (LPA), and moderate-to-vigorous PA (MVPA) were assessed objectively for 41 Flemish men using the SenseWear Pro 3 Armband. CpG site-specific methylation in leukocytes was determined using the Illumina HumanMethylation 450 BeadChip. Correlations were calculated between time spent on the three PA intensity levels and global DNA methylation, using a z-score-based method to determine global DNA methylation levels. To determine whether CpG site-specific methylation can be predicted by these three PA intensity levels, linear regression analyses were performed. Based on the significantly associated CpG sites at α = 0.005, lists were created including all genes with a promoter region overlapping these CpG sites. A biological pathway analysis determined to what extent these genes are overrepresented within several pathways. No significant associations were observed between global DNA methylation and SB (r = 0.084), LPA (r = -0.168), or MVPA (r = -0.125), although the direction of the correlation coefficients is opposite to what is generally reported in literature. SB has a different impact on global and gene-specific methylation than PA, but also LPA and MVPA affect separate genes and pathways. Furthermore, the function of a pathway seems to determine its association with SB, LPA, or MVPA. Multiple PA intensity levels, including SB, should be taken into account in future studies investigating the effect of physical (in)activity on human health through epigenetic mechanisms.

Limited potential of genetic predisposition scores to predict muscle mass and strength performance in Flemish Caucasians between 19 and 73 years of age.

Since both muscle mass and strength performance are polygenic in nature, the current study compared four genetic predisposition scores (GPS) in their ability to predict these phenotypes. Data were gathered within the framework of the first-generation Flemish Policy Research Centre "Sport, Physical Activity and Health" (2002-2004). Results are based on muscle characteristics data of 565 Flemish Caucasians (19-73 yr, 365 men). Skeletal muscle mass was determined from bioelectrical impedance. The Biodex dynamometer was used to measure isometric (PTstatic120°) and isokinetic strength (PTdynamic60° and PTdynamic240°), ballistic movement speed (S20%), and muscular endurance (Work) of the knee extensors. Genotyping was done for 153 gene variants, selected on the basis of a literature search and the expression quantitative trait loci of selected genes. Four GPS were designed: a total GPS (based on the sum of all 153 variants, each favorable allele = score 1), a data-driven and weighted GPS [respectively, the sum of favorable alleles of those variants with significant b-coefficients in stepwise regression (GPSdd), and the sum of these variants weighted with their respective partial r2 (GPSw)], and an elastic net GPS (based on the variants that were selected by an elastic net regularization; GPSen). It was found that four different models for a GPS were able to significantly predict up to ~7% of the variance in strength performance. GPSen made the best prediction of SMM and Work. However, this was not the case for the remaining strength performance parameters, where best predictions were made by GPSdd and GPSw.

In silico detection of phylogenetic informative Y-chromosomal single nucleotide polymorphisms from whole genome sequencing data.

A state-of-the-art phylogeny of the human Y-chromosome is an essential tool for forensic genetics. The explosion of whole genome sequencing (WGS) data due to the rapid progress of next-generation sequencing facilities is useful to optimize and to increase the resolution of the phylogenetic Y-chromosomal tree. The most interesting Y-chromosomal variants to increase the phylogeny are SNPs (Y-SNPs) especially since the software to call them in WGS data and to genotype them in forensic assays has been optimized over the past years. The PENNY software presented here detects potentially phylogenetic interesting Y-SNPs in silico based on SNP calling data files and classifies them into different types according to their position in the currently used Y-chromosomal tree. The software utilized 790 available male WGS samples of which 172 had a high SNP calling quality. In total, 1269 Y-SNPs potentially capable of increasing the resolution of the Y-chromosomal phylogenetic tree were detected based on a first run with PENNY. Based on a test panel of 57 high-quality and 618 low-quality WGS samples, we could prove that these newly added Y-SNPs indeed increased the resolution of the phylogenetic Y-chromosomal analysis substantially. Finally, we performed a second run with PENNY whereby all samples including those of the test panel are used and this resulted in 509 additional phylogenetic promising Y-SNPs. By including these additional Y-SNPs, a final update of the present phylogenetic Y-chromosomal tree which is useful for forensic applications was generated. In order to find more convincing forensic interesting Y-SNPs with this PENNY software, the number of samples and variety of the haplogroups to which these samples belong needs to increase. The PENNY software (inclusive the user manual) is freely available on the website http://bio.kuleuven.be/eeb/lbeg/software.