New variable selection strategy for analysis of high-dimensional DNA methylation data

In genetic association studies, regularization methods are often used due to their computational efficiency for analysis of high-dimensional genomic data. DNA methylation data generated from Infinium HumanMethylation450 BeadChip Kit have a group structure where an individual gene consists of multiple Cytosine-phosphate-Guanine (CpG) sites. Consequently, group-based regularization can precisely detect outcome-related CpG sites. Representative examples are sparse group lasso (SGL) and network-based regularization. The former is powerful when most of the CpG sites within the same gene are associated with a phenotype outcome. In contrast, the latter is preferred when only a few of the CpG sites within the same gene are related to the outcome. In this paper, we propose new variable selection strategy based on a selection probability that measures selection frequency of individual variables selected by both SGL and network-based regularization. In extensive simulation study, we demonstrated that the proposed strategy can show relatively outstanding selection performance under any situation, compared with both SGL and network-based regularization. Also, we applied the proposed strategy to identify differentially methylated CpG sites and their corresponding genes from ovarian cancer data.

The parvalbumin-positive interneurons in the mouse dentate gyrus express GABAA receptor subunits α1, β2, and δ along their extrasynaptic cell membrane

Neuronal circuitries in the hippocampus are involved in navigation and memory and are controlled by major networks of GABAergic interneurons. Parvalbumin (PV)-expressing interneurons in the dentate gyrus (DG) are identified as fast-spiking cells, playing a crucial role in network oscillation and synchrony. The inhibitory modulation of these interneurons is thought to be mediated mainly through GABAA receptors, the major inhibitory neurotransmitter receptors in the brain. Here we show that all PV-positive interneurons in the granular/subgranular layer (GL/SGL) of the mouse DG express high levels of the GABAA receptor δ subunit. PV-containing interneurons in the hilus and the molecular layer, however, express the δ subunit to a lower extent. Only 8% of the somatostatin-containing interneurons express the δ subunit, whereas calbindin- or calretinin-containing interneurons in the DG seem not to express the GABAA receptor δ subunit at all. Hence, these cells receive a GABAergic control different from that of PV-containing interneurons in the GL/SGL. Experiments investigating a possible co-expression of GABAA receptor α1, α2, α3, α4, α5, β1, β2, β3, or γ2 subunits with PV and δ subunits indicated that α1 and β2 subunits are co-expressed with δ subunits along the extrasynaptic membranes of PV-interneurons. These results suggest a robust tonic GABAergic control of PV-containing interneurons in the GL/SGL of the DG via δ subunit-containing receptors. Our data are important for better understanding of the neuronal circuitries in the DG and the role of specific cell types under pathological conditions.