General and specific combining abilities in a maize (Zea mays L.) test-cross hybrid panel: relative importance of population structure and genetic divergence between parents.

KEY MESSAGE: General and specific combining abilities of maize hybrids between 288 inbred lines and three tester lines were highly related to population structure and genetic distance inferred from SNP data. Many studies have attempted to provide reliable and quick methods to identify promising parental lines and combinations in hybrid breeding programs. Since the 1950s, maize germplasm has been organized into heterotic groups to facilitate the exploitation of heterosis. Molecular markers have proven efficient tools to address the organization of genetic diversity and the relationship between lines or populations. The aim of the present work was to investigate to what extent marker-based evaluations of population structure and genetic distance may account for general (GCA) and specific (SCA) combining ability components in a population composed of 800 inter and intra-heterotic group hybrids obtained by crossing 288 inbred lines and three testers. Our results illustrate a strong effect of groups identified by population structure analysis on both GCA and SCA components. Including genetic distance between parental lines of hybrids in the model leads to a significant decrease of SCA variance component and an increase in GCA variance component for all the traits. The latter suggests that this approach can be efficient to better estimate the potential combining ability of inbred lines when crossed with unrelated lines, and limits the consequences of tester choice. Significant residual GCA and SCA variance components of models taking into account structure and/or genetic distance highlight the variation available for breeding programs within structure groups.

Characterizing individual differences in animal responses to a nutritional challenge: Toward improved robustness measures.

The aim of this paper was to explore the variation between individuals in the response to and recovery from a nutritional challenge, the repeatability of responses between lactation stages, and the use of shape-clustering methods to classify similar individuals. Sixteen dairy goats were exposed to a 2-d nutritional challenge (underfeeding) at 2 different stages of lactation. Each challenge consisted of a 7-d control period with standard total mixed ration (TMR), 2d of straw-only feeding, and a 10-d recovery period on the TMR. All feeds were offered ad libitum, as was water. The first challenge was in late lactation on primiparous goats (mean days in milk=249), and the second challenge was carried out on the same goats early in the following lactation (mean days in milk=28). The main energetic response traits dry matter intake (DMI), milk yield, body weight, milk fat and protein contents, and plasma glucose, fatty acids (NEFA), ß-hydroxybutyrate (BHB), urea, and insulin, were measured daily throughout. A clustering procedure linked to a piecewise mixed model was used to characterize different types of response. As expected, straw feeding caused a large decline in DMI and milk yield, and substantial increases in milk fat and milk protein composition, relative to the prechallenge period on the control TMR. For both DMI and milk yield, the slope of the response, and hence the size of the drop, was strongly related to the prechallenge values, indicating that these 2 measures were tightly constrained by the challenge. Regression slopes between lactation stages for responses to the same nutritional challenge were significant for DMI, milk protein content, plasma BHB and urea, and body weight, indicating that within-animal responses in late and early lactation were repeatable. The clustering procedure generally performed well, classifying both scaling differences and differences in shape. The extent of reranking of cluster designations between late lactation and the following early lactation period was examined. For milk yield, DMI, body weight, and urea, relatively little reranking occurred; the numbers of goats not changing class number were 10, 12, 10, and 13, respectively. In contrast, for milk contents of fat and protein, and also for BHB, no clear association was found between late and early lactation class numbers. For NEFA and glucose, these comparisons were not relevant because either the vast majority of goats were in 1 cluster (NEFA) or because an outlier goat skewed the cluster designation (glucose in late lactation). For insulin, 9 out of 16 goats kept the same rank.

Two-dimensional segmentation for analyzing Hi-C data.

MOTIVATION: The spatial conformation of the chromosome has a deep influence on gene regulation and expression. Hi-C technology allows the evaluation of the spatial proximity between any pair of loci along the genome. It results in a data matrix where blocks corresponding to (self-)interacting regions appear. The delimitation of such blocks is critical to better understand the spatial organization of the chromatin. From a computational point of view, it results in a 2D segmentation problem. RESULTS: We focus on the detection of cis-interacting regions, which appear to be prominent in observed data. We define a block-wise segmentation model for the detection of such regions. We prove that the maximization of the likelihood with respect to the block boundaries can be rephrased in terms of a 1D segmentation problem, for which the standard dynamic programming applies. The performance of the proposed methods is assessed by a simulation study on both synthetic and resampled data. A comparative study on public data shows good concordance with biologically confirmed regions. AVAILABILITY AND IMPLEMENTATION: The HiCseg R package is available from the Comprehensive R Archive Network and from the Web page of the corresponding author.