Single Marker Family-Based Association Analysis Not Conditional on Parental Information.

Whole genome prediction (WGP) modeling and genome-wide association (GWA) analyses are big data issues in agricultural quantitative genetics. Both areas require meaningful input from the statistical scholarly community in order to further improve the accuracy of prediction of genetic merit and inference on putative causal variants as well as improving the computational efficiency of existing methods and algorithms. These concerns have become increasingly critical as new sequencing technologies will only exacerbate current model dimensionality problems. We focus primarily on mixed model and hierarchical Bayesian analyses which have been most commonly pursued by animal and plant breeders for WGP thus far. We draw attention to our observation that many such previous analyses have not carefully inferred upon hyperparameters defined at the top levels of the Bayesian model hierarchy, but simply arbitrarily specify their values. We also reassess previous discussions on WGP model dimensionality, believing that useful data augmentation schemes utilized in various Markov Chain Monte Carlo (MCMC) schemes have led to a general misunderstanding that heavy-tailed or variable selection-based WGP models may be highly parameterized relative to more standard mixed model representations. Computational efficiency is addressed with respect to MCMC and competitive, albeit approximate, alternatives. Furthermore, GWA analyses are reassessed, encouraging a greater reliance on shrinkage-based inferences based on critically chosen priors, instead of potentially nonreproducible fixed effects P value-based inference.

Full text Figures and data Side by side Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Appropriate modulation of escape behaviors in response to potentially damaging stimuli is essential for survival. Although nociceptive circuitry has been studied, it is poorly understood how genetic contexts affect relevant escape responses. Using an unbiased genome-wide association analysis, we identified an Ly6/α-neurotoxin family protein, Belly roll (Bero), which negatively regulates Drosophila nociceptive escape behavior. We show that Bero is expressed in abdominal leucokinin-producing neurons (ABLK neurons) and bero knockdown in ABLK neurons resulted in enhanced escape behavior. Furthermore, we demonstrated that ABLK neurons responded to activation of nociceptors and initiated the behavior. Notably, bero knockdown reduced persistent neuronal activity and increased evoked nociceptive responses in ABLK neurons. Our findings reveal that Bero modulates an escape response by regulating distinct neuronal activities in ABLK neurons. Editor's evaluation This convincing and fundamental study uses unbiased genome-wide association analysis to identify a gene, called Bero, encoding Ly6/α-neurotoxin family protein, that affects the way larval Drosophila respond to nociceptive stimuli. This discovery is followed up by the identification of neurons in which Bero function is relevant for the modulation of nociceptive behaviour, namely the abdominal leucokinin-producing neurons. These neurons are activated by nociceptive sensory neurons and can initiate escape behavior. In these neurons Bero modulates both persistent and evoked activities. This elegant work will be of interest to neurobiologists working on genes, neural circuits, and behaviour. https://doi.org/10.7554/eLife.83856.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Appropriate escape behavior in response to threatening stimuli is essential for organismal survival (Branco and Redgrave, 2020; Burrell, 2017; Chin and Tracey, 2017; Im and Galko, 2012; Peirs and Seal, 2016). In both vertebrates and invertebrates, the modulation of escape behavior depends on environmental contexts as well as intensity and modality of stimuli, which are sensed and processed in a multilayered neuronal network, subject to genetic variation, that integrates external and internal sensory stimuli (Burnett et al., 2016; Jennings et al., 2014; Mu et al., 2012; Ohyama et al., 2015). Within the network, peptidergic neuromodulation plays a critical role in neuronal information processing (Nässel and Winther, 2010; Taghert and Nitabach, 2012). When Drosophila melanogaster larvae encounter noxious stimuli, such as harsh mechanical stimulations intrinsic to attacks by parasitoid wasps, they show stereotypic escape behaviors composed of an initial abrupt bending and subsequent corkscrew-like rolling (Hwang et al., 2007; Onodera et al., 2017; Tracey et al., 2003). In larvae, the noxious stimuli are mainly sensed by class IV dendritic arborization neurons (Class IV neurons) underneath the body wall (Tracey et al., 2003). In addition, a type of tracheal dendrite neuron (v’td2 neurons) and class III dendritic arborization neurons (Class III neurons) are known to detect noxious light stimuli and noxious cold stimuli, respectively (Imambocus et al., 2022; Turner et al., 2016). Recent studies have identified several important downstream components of the nociceptive neural circuitry (Ohyama et al., 2015; Takagi et al., 2017; Yoshino et al., 2017). In parallel, several genes and neurotransmitters have been characterized as context-dependent modulators of the behaviors (Burgos et al., 2018; Dason et al., 2020; Hu et al., 2017; Kaneko et al., 2017). While most of these studies identified the nociceptive neural circuitry by connectomic approaches, we used a genetic strategy to provide additional insights into the nociceptive modulation. Importantly, we discovered that D. melanogaster larvae displayed an extensive natural variation in nociceptive escape behavior. This broad behavioral variation allowed us to perform an unbiased genome-wide association (GWA) analysis to explore genetic variants affecting the nociceptive escape behaviors. In this study, we have demonstrated that an Ly6/α-neurotoxin family protein, Belly roll (Bero), modulates the neuronal excitability of peptidergic interneurons that initiates and facilitates Drosophila nociceptive escape behaviors. We first identified the bero gene through GWA analysis and validated its role in regulating nociception using pan-neuronal bero knockdown and null mutant animals. By searching for bero-expressing neurons, we identified abdominal leucokinin-producing neurons (ABLK neurons) that have been reported to be involved in controlling larval escape behaviors (Hu et al., 2020; Imambocus et al., 2022). We then found that bero knockdown in ABLK neurons resulted in enhanced escape behaviors, similar to bero mutants. Furthermore, we demonstrated that ABLK neurons responded to the activation of nociceptors and initiated nociceptive escape behavior. Notably, bero knockdown in ABLK neurons resulted in reduced persistent activities and increased evoked nociceptive responses in ABLK neurons, implying an essential role of bero in regulating neuronal activities. Based on our results regarding the ABLK neurons, we propose that these neurons may integrate external and internal sensory stimuli, thereby orchestrating nociceptive escape behaviors. Results Natural variation in nociceptive responses in wild-type strains of the Drosophila melanogaster Genetic Reference Panel Upon noxious thermal stimulation, Drosophila larvae show stereotypical nociceptive responses: bending and rolling behavior (Figure 1A; Tracey et al., 2003). Using the Heat Probe Assay (HPA; Tracey et al., 2003), we noticed that there is a large difference in the latency of rolling behavior (rolling latency) between the two commonly used strains, w1118 and Canton-S (Figure 1B). In order to evaluate natural variation in nociceptive responses, we took advantage of the Drosophila melanogaster Genetic Reference Panel (DGRP; Huang et al., 2014; Mackay et al., 2012), a collection of wild-caught Drosophila lines. We quantified the rolling behavior in 38 representative inbred wild-type strains of DGRP and observed a considerable variation in the behavior: the strain-specific median rolling latency ranged from 2.83 s to 10 s and the occurrence rates of rolling behavior (rolling probability) within 10 s ranged from around 30% to almost 100% among these strains (Figure 1C and D). We categorized the rolling probability into three response classes (rolling probability in 2, 5, and 10 s) based on previous studies (Onodera et al., 2017; Terada et al., 2016; Tracey et al., 2003) and found that the three response classes are highly correlated with one another (Figure 1—figure supplement 1A). To identify genetic variants associated with the behavioral variation, we performed multiple trials of GWA mapping analyses using the publicly available web-based analysis tool, DGRP 2.0 (Huang et al., 2014), with four statistical metrics that characterize larval rolling behavior and are mutually dependent to some extent (Figure 1E, Figure 1—source data 1, ‘Materials and methods’). We then identified in total 37 single-nucleotide polymorphisms (SNPs), two deletions, and three insertions in or near 36 candidate genes (Figure 1—source data 2), at a threshold p-value of 1.0 × 10–5, a commonly used threshold p-value in previous studies with DGRP (Figure 1F, Figure 1—figure supplement 1B; Mackay and Huang, 2018). Figure 1 with 1 supplement see all Download asset Open asset Natural diversity in nociceptive rolling escape behavior in wild-type strains of Drosophila melanogaster. (A) A schematic representation of the Heat Probe Assay (see ‘Materials and methods’ for details). (B) The latency of rolling escape behavior (rolling latency) is different between two wild-type strains, w1118 and Canton-S (CS) (46°C, [w1118] n = 40, [CS] n = 38; 48°C, [w1118] n = 32, [CS] n = 33; 50°C, [w1118] n = 33, [CS] n = 33; ***p<0.001, Wilcoxon rank-sum test). Violin plots provide a kernel density estimate of the data, where the middle circle shows the median, a boxplot shape indicates 25th and 75th percentiles, and whiskers to the left and right of the box indicate the 90th and 10th percentiles, respectively, in this and the following figures. Each data point represents an individual larva. See source data tables for detailed genotypes in this and the following figures. (C) The rolling latency of 38 representative Drosophila melanogaster Genetic Reference Panel (DGRP) lines in ascending order (n = 30 or 31 larvae/line). Boxplots indicate the median and 25th and 75th percentiles, and whiskers to the left and right of the box indicate the 90th and 10th percentiles, respectively, in this and the following figures. (D) The rolling probability of 38 representative DGRP lines in ascending order (n = 30 or 31 larvae/line). The stacked bar chart indicates the rolling probability within 2, 5, and 10 s, respectively. (E) A diagram showing experimental procedures for the genome-wide association (GWA) analysis. (F) GWA analysis for median rolling latency of the rolling escape behavior. The p-values (−log10 transformed) are shown on the y axis. The gray dotted line marks the nominal p-value threshold (1.0 × 10–5). Each data point corresponds to an individual genetic variant (single-nucleotide polymorphism, deletion, or insertion). Data points are arranged by relative chromosome (genomic) position, the color code indicates the respective chromosome to which they belong. See also Figure 1—source data 1 and Figure 1—figure supplement 1A. Figure 1—source data 1 Summary table of genotypes, statistical testing, and graph data for Figure 1. https://cdn.elifesciences.org/articles/83856/elife-83856-fig1-data1-v1.zip Download elife-83856-fig1-data1-v1.zip Figure 1—source data 2 Genome-wide association results for rolling behavior. https://cdn.elifesciences.org/articles/83856/elife-83856-fig1-data2-v1.zip Download elife-83856-fig1-data2-v1.zip Identification of CG9336/belly roll (bero) through an RNA interference knockdown screen of candidate genes To explore impacts of these candidate genes on nociceptive responses, we carried out a secondary functional screen by pan-neuronal RNA interference (RNAi)-mediated gene knockdown. By evaluating 17 RNAi strains that were currently available, we found that pan-neuronal knockdown of the CG9336 gene resulted in significant enhancement of the rolling probability (Figure 2A). To further test whether the phenotype was specifically caused by suppression of CG9336, we generated two distinct short-hairpin RNAs (shRNAs) against CG9336 (shRNA#1 and #2; Figure 2B; ‘Materials and methods’), validated the efficiency of shRNA#2 (Figure 2—figure supplement 1A and B; ‘Materials and methods’), and found that both shRNAs caused a comparable enhancement (Figure 2E, Figure 2—figure supplement 1C). We also performed three sets of effector-control tests for UAS-bero shRNA#2 with different genomic backgrounds. In each group, both the UAS-bero shRNA#2 control animal and the corresponding controls exhibited comparable rolling behavior, which ensures that any observed phenotypes were not due to leaky expression from the UAS transgene (Figure 2E). We then designated the gene as belly roll (bero), after a high-jump style in athletics. bero encodes a member of the Ly6/α-neurotoxin protein superfamily, which is characterized by a three-finger structure, including various membrane-tethered and secreted polypeptides (Loughner et al., 2016). Using several bioinformatics tools, we predicted a signal peptide sequence at the amino terminus, two possible glycosylation sites, a GPI-anchorage site, a putative transmembrane region at the carboxyl terminus, and five putative disulfide bridges, which are typical structural features of the Ly6/α-neurotoxin protein superfamily (Figure 2C, Figure 2—figure supplement 1D; Almagro Armenteros et al., 2019; Baek et al., 2021; Eisenhaber et al., 1999; Gupta and Brunak, 2001; Krogh et al., 2001). A three-dimensional model of Bero was generated by protein structure prediction using AlphaFold2 (Jumper et al., 2021; Figure 2D). Figure 2 with 2 supplements see all Download asset Open asset bero negatively regulates nociceptive rolling escape behavior. (A) A secondary functional screen by pan-neuronal RNA interference. The rolling latency of each UAS-RNAi line was measured. Pan-neuronal knockdown of CG9336/bero reduces rolling latency (Heat Probe Assay, n = 30 or 31 larvae/genotype; ****p<0.0001, Wilcoxon rank-sum test). (B) A schematic representation of bero-associated SNP (2L_20859945_SNP; minor allele, cytosine; major allele, thymine; minor allele frequency = 0.4167). (C) A schematic representation of unprocessed and mature Bero protein. SP, signal peptide; TM, transmembrane region. Orange lines mark the predicted disulfide bonds. Yellow lines mark the predicted N-glycosylation sites. Pink lines mark the predicted GPI-modification site. See also Figure 2—figure supplement 1D. (D) A three-dimensional protein structure prediction of unprocessed Bero protein by AlphaFold2. Red region, signal peptide; blue region, transmembrane region; orange sticks, predicted disulfide bonds; yellow region, predicted N-glycosylation sites; pink region, predicted GPI-modification site. (E) Rolling latency of nSyb>attP2 control (n = 63), pan-neuronal bero knockdown animals (nSyb>bero RNAishRNA#2, n = 65), and the corresponding three sets of effector-controls (w1118 background: attP2 control, n = 64, bero RNAishRNA#2, n = 60; Canton-S background: attP2 control, n = 60, bero RNAishRNA#2, n = 60, y w background: attP2 control, n = 55, bero RNAishRNA#2, n = 55). **p<0.01; n.s., nonsignificant; Wilcoxon rank-sum test. The genetic backgrounds are controlled (see source data tables for detailed genotypes in this and the following figures). (F) Rolling latency of Canton-S control (n = 38), bero heterozygous (beroKO/+, n = 38), and homozygous (beroKO/beroKO, n = 37) mutant animals. **p<0.01, ***p<0.001, Wilcoxon rank-sum test. The bero KO strain had been outcrossed to Canton-S for 11 generations. Figure 2—source data 1 Summary table of genotypes, statistical testing, and graph data for Figure 2. https://cdn.elifesciences.org/articles/83856/elife-83856-fig2-data1-v1.zip Download elife-83856-fig2-data1-v1.zip The bero-associated SNP, 2L_20859945_SNP, is located in the second intron of CG9336 isoform A (FBtr0081427) and comprises two nucleotide variations in the DGRP core 40 lines: a minor allele, cytosine, and a major allele, thymine (Figure 2B). We found that most of the 38 DGRP core lines with the bero minor allele (C) showed higher responsiveness than those with the major allele (T) upon noxious thermal stimulations (Figure 2—figure supplement 2A). Based on these observations and our bero knockdown results, we speculated that animals with the bero minor allele (C) might have lower expression levels of bero compared with those possessing the major allele (T). Thus, using reverse transcriptase PCR, we compared the expression levels of bero in the larval central nervous system (CNS) among four DGPR strains that showed extremely quick or slow behavioral responses. We found that animals with the minor allele have lower expression levels of bero (Figure 2—figure supplement 2B). Using online sequence analysis based on JASPAR 2022 (Castro-Mondragon et al., 2022), we predicted that the neighboring nucleotide sequences of the bero-associated SNP might be a target site of certain transcription factors that control the expression level of bero in CNS (Supplementary file 2). Furthermore, we generated a bero null mutant line using CRISPR-mediated mutagenesis, which deleted the entire protein-coding sequence (Figure 2—figure supplement 1E; ‘Materials and methods’), and observed in HPA experiments that both bero heterozygous (beroKO/+) and homozygous (beroKO/beroKO) mutant animals displayed an enhanced rolling escape behavior (Figure 2F). Taken together, these results strongly suggest that bero expression in neurons negatively regulates nociceptive escape behavior. bero is expressed in several subgroups of peptidergic neurons in the larval CNS We next sought to find which neurons express bero in the larval CNS. To do this, we examined the overlap between bero-expressing neurons labeled in a protein-trap strain bero-YFP (CPTI-001654; Lowe et al., 2014), in which the endogenous Bero protein is fused with a yellow fluorescent protein variant mVenus, and candidate neurons labeled by a CD4:tdTomato fusion protein driven by distinct GAL4-drivers. We found no bero-YFP expression in the Class IV neurons (nociceptors) or in the downstream neurons that have been reported to be required for the nociceptive escape behavior (Figure 3—figure supplement 1A; Ohyama et al., 2015). Consistent with this, bero knockdown specifically in the corresponding neurons did not alter the rolling behavior (Figure 3—figure supplement 1B). Instead, we observed bero-YFP expression in several groups of peptidergic neurons specified by the bHLH transcription factor dimmed (dimm), including insulin-producing cells (IPCs) and a subset of Eclosion hormone-producing neurons (EH neurons) in the larval brain, and abdominal leucokinin-producing neurons (ABLK neurons) in the lateral region of abdominal ganglions in the larval ventral nerve cord (VNC; Figure 3A and B). In support of our observations, a recent single-cell transcriptomic atlas showed that bero/CG9336 is selectively enriched in peptidergic neurons (Corrales et al., 2022). Bero is also expressed in certain non-neuronal tissues, such as midline glia (Figure 3A). However, we did not study Bero function in the glia in this study. Figure 3 with 1 supplement see all Download asset Open asset Expression of bero in ABLK neurons plays an essential role in the negative regulation of nociceptive behavior. (A) Confocal image stacks (maximum projection) showing fluorescence of the endogenous Bero reporter, Bero-YFP (green), and the colocalized peptidergic neurons with GAL4, UAS-CD4-tdTomato (magenta) in third-instar larvae. Scale bars, 50 μm. (B) A schematic representation of bero-expressing neurons. IPC, insulin-producing cells; EH, eclosion hormone-producing neurons; ABLK, abdominal leucokinin-producing neurons; SEG, the subesophageal ganglion; VNC, the ventral nerve cord. (C) Confocal image stack showing endogenous Bero expression (labeled by Bero-YFP, yellow), LK neurons (Lk-GAL4.TH, UAS-CD4-tdTomato, magenta), and leucokinin (labeled by anti-Lk, cyan) in a third-instar larva. Bottom and left panels show an XZ and YZ cross-section of the ABLK somatic region (locations of the cross-sections are indicated by horizontal and vertical gray lines in the primary image), respectively. Middle panels show magnified views of the boxed regions: Individual optical sections depicting ABLK neurons (ABLK neurons are indicated by A1–A7 label, and an image of A7 with enhanced intensity is showed), Right: image stack depicting SELK neurons (indicated by arrowheads), and image stack depicting ABLK neurites (an image with enhanced intensity is showed). Scale bars: 50 μm; 10 μm or 20 μm for the magnified view. (D) Confocal image stack showing the dendrite (labeled by ABLK >DenMark, magenta) and axon terminal markers (labeled by ABLK >syt:eGFP, green) in ABLK neurons in a third-instar larva. Scale bars: 50 μm; 20 μm for the magnified view. (E) Rolling latency of LK-specific bero knockdown larvae (Lk>bero RNAishRNA #2, n = 73) and control (Lk>attP2, n = 75). Rolling latency of Eh-specific bero knockdown larvae (Eh>bero RNAishRNA #2, n = 38) and control (Eh>attP2, n = 41). Rolling latency of Ilp2-specific bero knockdown larvae (Ilp2>bero RNAishRNA #2, n = 40) and control (Ilp2>attP2, n = 39). ***p<0.001; n.s., nonsignificant; Wilcoxon rank-sum test. (F) Rolling latency of ABLK-specific bero overexpression larvae (ABLK>Bero:FLAG, n = 72), control (ABLK>myr:GFP, n = 73), and the corresponding effector-controls (y v background: myr:GFP control, n = 60, Bero:FLAG, n = 61). **p<0.01; n.s., nonsignificant; Wilcoxon rank-sum test. Figure 3—source data 1 Summary table of genotypes, statistical testing, and graph data for Figure 3. https://cdn.elifesciences.org/articles/83856/elife-83856-fig3-data1-v1.zip Download elife-83856-fig3-data1-v1.zip bero expression in ABLK neurons negatively regulates nociceptive rolling escape behavior To identify the particular subgroups of neurons that require bero expression for negative regulation of nociceptive escape behavior, we examined rolling escape behavior in larvae with cell type-specific RNAi knockdowns of bero. We found that knockdown of bero using Lk-GAL4, a leucokinin-producing neuron-specific driver, caused enhanced rolling escape behavior (Figure 3E). In contrast, knockdown of bero using either Eh-GAL4, a neuron-specific driver, or an driver, did not in any in rolling escape behavior (Figure 3E). LK neurons can be into four distinct subgroups in the larval LK neurons lateral LK neurons subesophageal LK neurons and abdominal LK neurons et al., bero is expressed in all of ABLK neurons and is to not expressed in any LK neurons (Figure 3A and Consistent with this, the of bero-YFP labeled neurons were for (Figure of individual optical sections showed that Bero on both the and internal of ABLK neurons (Figure We expressed the dendrite and axon terminal namely and in ABLK neurons (Figure that Bero both and of the neurons. To the role of bero expression in ABLK neurons in rolling behavior, we to ABLK-specific bero overexpression using an et al., 2010; 2016; Figure 3—figure supplement 1C). We observed that bero overexpression rolling behavior (Figure that bero expression in ABLK neurons plays an essential role in the negative regulation of nociceptive behavior. In contrast, in the effector-control both the control animal and the corresponding controls showed comparable levels of rolling behavior (Figure we our analysis on the of ABLK neurons and the functional of bero in these neurons. bero persistent activities and evoked nociceptive responses in ABLK neurons We next whether ABLK neurons respond to of To test this, we expressed a blue variant of et al., in and a fluorescence et al., in LK neurons. We responses in ABLK neurons in and observed that cell and of ABLK neurons displayed persistent of (Figure and Figure supplement which was with a previous study et al., Notably, activation of the nociceptors evoked evoked neuronal activities in ABLK neurons, that these neurons respond to nociceptive (Figure and Hu et al., We used larvae on a with no additional all as controls activation in the A was also in the control larvae on a with no additional the the of the blue light (Figure This indicates that the be to the persistent in ABLK neurons. Figure with 2 supplements see all Download asset Open asset bero is for the of persistent activities and suppression of evoked nociceptive activity in ABLK neurons. (A) results of responses showing the persistent signal of ABLK neurons in control and bero knockdown larvae bero RNAishRNA (B) A of the of persistent activities of ABLK neurons in control (n = neurons from three and bero knockdown larvae (n = neurons from three ****p<0.0001, Wilcoxon rank-sum test. (C) of the nociceptive responses of ABLK neurons in control and bero knockdown larvae bero RNAishRNA after activation of Class IV neurons Scale bars, 10 μm. (D) for responses see ‘Materials and methods’ for of ABLK neurons in control n = neurons from three and bero knockdown larvae bero RNAishRNA #2, n = neurons from three upon activation of Class IV neurons light is indicated by at and for blue and gray indicate (E) for responses of ABLK neurons in control n = neurons from three and bero knockdown larvae bero RNAishRNA #2, n = neurons from three upon activation of Class IV neurons. (F) of responses of ABLK neurons in control n = neurons from three n = neurons from three and bero knockdown larvae n = neurons from three n = neurons from three upon activation of Class IV neurons n.s., nonsignificant; test followed by test. Figure data 1 Summary table of genotypes, statistical testing, and graph data for Figure Download To further the of bero in ABLK neurons, we whether knockdown of bero persistent activities nociceptive responses of the neurons. We found that bero knockdown almost persistent activities (Figure and it increased the of evoked nociceptive responses (Figure and that bero knockdown enhanced rolling escape behavior in the HPA experiments (Figure we speculated that the evoked nociceptive responses in ABLK neurons are important for rolling escape behavior. the phenotype of the bero knockdown on persistent activities in ABLK neurons, we performed a behavioral analysis of and found that bero knockdown did not affect larval (Figure supplement persistent in ABLK neurons of bero knockdown animals to be enhanced compared to controls (Figure Figure supplement 1B). We that the used for in nociceptive Furthermore, ABLK neurons with bero knockdown displayed increased to these nociceptive than which may provide a possible the neural activity may not Thus, we compared activity the we have that bero plays a critical role in not a higher level of persistent also the evoked nociceptive responses of the ABLK neurons, thereby the nociceptive escape behavior (Figure of evoked activity in ABLK neurons is for the control of rolling escape behavior To test whether of ABLK activity to escape behaviors, we activated the neurons using a et al., We first activated neurons and found that light a rolling behavior that was observed in control larvae (Figure and B). We then the to selectively express in ABLK or SELK neurons (Figure activation of ABLK neurons almost resulted in not rolling bending behavior, which is as a initial of the entire sequence of escape (Figure Figure supplement 1A and B). In contrast, activation of SELK neurons rolling behavior in all larvae. Figure with 1 supplement see all

Information from multiple informants is frequently used to assess psychopathology. We consider marginal regression models with multiple informants as discrete predictors and a time to event outcome. We fit these models to data from the Stirling County Study; specifically, the models predict mortality from self report of psychiatric disorders and also predict mortality from physician report of psychiatric disorders. Previously, Horton et al. found little relationship between self and physician reports of psychopathology, but that the relationship of self report of psychopathology with mortality was similar to that of physician report of psychopathology with mortality. Generalized estimating equations (GEE) have been used to fit marginal models with multiple informant covariates; here we develop a maximum likelihood (ML) approach and show how it relates to the GEE approach. In a simple setting using a saturated model, the ML approach can be constructed to provide estimates that match those found using GEE. We extend the ML technique to consider multiple informant predictors with missingness and compare the method to using inverse probability weighted (IPW) GEE. Our simulation study illustrates that IPW GEE loses little efficiency compared with ML in the presence of monotone missingness. Our example data has non-monotone missingness; in this case, ML offers a modest decrease in variance compared with IPW GEE, particularly for estimating covariates in the marginal models. In more general settings, e.g., categorical predictors and piecewise exponential models, the likelihood parameters from the ML technique do not have the same interpretation as the GEE. Thus, the GEE is recommended to fit marginal models for its flexibility, ease of interpretation and comparable efficiency to ML in the presence of missing data.

Appearance and milling quality are two crucial properties of rice grains affecting its market acceptability. Understanding the genetic base of rice grain quality could considerably improve the high quality breeding. Here, we carried out an association analysis to identify QTL affecting nine rice grain appearance and milling quality traits using a diverse panel of 258 accessions selected from 3K Rice Genome Project and evaluated in two environments Sanya and Shenzhen. Genome-wide association analyses using 22,488 high quality SNPs identified 72 QTL affecting the nine traits. Combined gene-based association and haplotype analyses plus functional annotation allowed us to shortlist 19 candidate genes for seven important QTL regions affecting the grain quality traits, including two cloned genes (GS3 and TUD), two fine mapped QTL (qGRL7.1 and qPGWC7) and three newly identified QTL (qGL3.4, qGW1.1, and qGW10.2). The most likely candidate gene(s) for each important QTL were also discussed. This research demonstrated the superior power to shortlist candidate genes affecting complex phenotypes by the strategy of combined GWAS, gene-based association and haplotype analyses. The identified candidate genes provided valuable sources for future functional characterization and genetic improvement of rice appearance and milling quality.

Measures of cardiac structure such as left ventricular mass (LVM) have been shown to be significantly heritable. Relative to other complex traits, however, gene variants influencing echocardiographic traits have been less frequently identified. In this study of subjects from the Fels Longitudinal Study, our aim was to perform quantitative trait linkage and genome-wide association (GWA) analyses on LVM, relative wall thickness (RWT), fractional shortening (FS), aortic root diameter (ARD), and LVM index (LVMI; LVM/height 2.7 ). The study sample consisted of 1,060 (467 male, 593 female) whites aged 8-97 years from 131 families. Variables were inverse normal transformed for all analyses. Each subject was genotyped with the Illumina Human 610-Quad BeadChip containing more than 550,000 single nucleotide polymorphism (SNP) markers. Variance-component linkage analyses were conducted using a subset of 17,561 genome-wide SNP markers in genetic equilibrium with each other. GWA analyses were conducted using all SNPs and the measured genotype method implemented in SOLAR, allowing for residual non-independence among relatives After adjusting for covariates of age, sex, BMI, and population stratification, all echocardiograpic traits were significantly heritable (p &lt; 0.00006); h 2 (SE) for LVM, LVMI, RWT, FS, and ARD were 0.28 (0.07), 0.36 (0.07), 0.22 (0.07), 0.27 (0.06), and 0.55 (0.06), respectively. Linkage analyses revealed significant linkage of LVMI to chromosome 1p31 (LOD=3.64 at 96cM) in a region where the LEPR (leptin receptor), PDE4B (phosphodiesterase 4B, cAMP-specific), NEGR1 (neuronal growth regulator 1), and TNNI3K (TNNI3 interacting kinase) genes are located. For ARD, we identified chromosomal region 13q14 (LOD = 3.88 at 45cM) which harbors the FOXO1 (forkhead box O1) and TSC22D1 (TSC22 domain family, member 1) genes. Suggestive (LOD &gt; 2.0) linkages were found on chromosomes 1, 2, 3, 4, 9 and 19 for the various traits. In GWA analyses, significant ( χ 2 &gt;27.85) association was found between LVMI and SNP rs7939215 on chromosome 11p15. For LVM we observed evidence of a suggestive association ( χ 2 &gt;22.70) with rs660464 in the NRG3 (neuregulin 3) gene, which has been associated with hypertrophy among the Amish. Other suggestive associations were found between FS and SNPs rs12485048 and rs5753622 on chromosome 22q12, and SNP rs2993639 chromosome on 3q21. In summary, we identified several significant linkage and association signals for cardiac measures. Future work will include replicating these results in other study populations.

Residual feed intake (RFI) is a complex trait that is economically important for livestock production; however, the genetic and biological mechanisms regulating RFI are largely unknown in pigs. Therefore, the study aimed to identify single nucleotide polymorphisms (SNPs), candidate genes and biological pathways involved in regulating RFI using Genome-wide association (GWA) and pathway analyses. A total of 596 Yorkshire boars with phenotypes for two different measures of RFI (RFI1 and 2) and 60k genotypic data was used. GWA analysis was performed using a univariate mixed model and 12 and 7 SNPs were found to be significantly associated with RFI1 and RFI2, respectively. Several genes such as xin actin-binding repeat-containing protein 2 (XIRP2),tetratricopeptide repeat domain 29 (TTC29),suppressor of glucose, autophagy associated 1 (SOGA1),MAS1,G-protein-coupled receptor (GPCR) kinase 5 (GRK5),prospero-homeobox protein 1 (PROX1),GPCR 155 (GPR155), and FYVE domain containing the 26 (ZFYVE26) were identified as putative candidates for RFI based on their genomic location in the vicinity of these SNPs. Genes located within 50 kbp of SNPs significantly associated with RFI and RFI2 (q-value ≤ 0.2) were subsequently used for pathway analyses. These analyses were performed by assigning genes to biological pathways and then testing the association of individual pathways with RFI using a Fisher’s exact test. Metabolic pathway was significantly associated with both RFIs. Other biological pathways regulating phagosome, tight junctions, olfactory transduction, and insulin secretion were significantly associated with both RFI traits when relaxed threshold for cut-off p-value was used (p ≤ 0.05). These results implied porcine RFI is regulated by multiple biological mechanisms, although the metabolic processes might be the most important. Olfactory transduction pathway controlling the perception of feed via smell, insulin pathway controlling food intake might be important pathways for RFI. Furthermore, our study revealed key genes and genetic variants that control feed efficiency that could potentially be useful for genetic selection of more feed efficient pigs.

BackgroundUtilizing Single Nucleotide Polymorphism (SNP) marker technology, a phylogenetic and agronomic trait network analysis was conducted on the collected hot pepper and sweet pepper germplasm resources, providing a theoretical basis for parental selection and new varieties.ResultsSpecific-locus amplified fragment sequencing (SLAF-seq) technology was employed for a genome-wide association study (GWAS) on 197 hot pepper and sweet pepper germplasm resources, generating 1404.88 Mb clean reads data with an average Q30 of 91.5% and mean GC content of 37.96%. Through sequencing data analysis, a total of 639,815 SLAF tags were obtained with an average sequencing depth of 12.16x, among which 86,381 were polymorphic SLAF tags, leading to the development of 18,145,155 SNP markers. The identified SNP markers were used for cluster analysis of the genetic structure and phylogenetic relationships of hot pepper and sweet pepper germplasm resources, dividing the 197 hot pepper and sweet pepper germplasm resources into 9 clusters. Additionally, a genome-wide association analysis was conducted on 25 agronomic traits of the 197 hot pepper and sweet pepper materials, yielding a substantial number of significantly associated SNP loci with agronomic traits. A correlation network analysis diagram was drawn among the various agronomic traits, preliminarily determining the relationships between the 25 agronomic characteristics of hot pepper and sweet pepper and the positions of 15 agronomic traits (p < 1.707 × 10−8) on the chromosomes were annotated, forming multi-trait aggregation regions.ConclusionsOur research reveals the genetic diversity, phylogenetic relationships, and population structure of 197 hot pepper and sweet pepper germplasm resources, providing a basis for germplasm identification, resource utilization, and breeding.

In this chapter we consider dependent data but move from the linear models of Chap.8 to general regression models. As in Chap.6, we consider generalized linear models (GLMs) and, more briefly, nonlinear models. We first give an outline of this chapter. In Sect.9.2 we describe three motivating datasets to which we return throughout the chapter. The GLMs discussed in Sect.6.3 can be extended to incorporate dependences in observations on the same unit; as with the linear model, an obvious way to carry out modeling in this case is to introduce unit-specific random effects. Within a GLM a natural approach is for these random effects to be included on the linear predictor scale. The resultant conditional models are known as generalized linear mixed models (GLMMs), and these are introduced in Sect.9.3. In Sects.9.4 and9.5 we describe likelihood and conditional likelihood methods of estimation, with Sect.9.6 devoted to a Bayesian treatment. Section9.7 illustrates some of the flexibility of GLMMs by describing and applying a particular model for spatial dependence. An alternative random effects specification, based on conjugacy, is described in Sect.9.8. An important approach to the modeling and analysis of dependent data that is philosophically different from the random effects formulation is via marginal models and generalized estimating equations (GEE), and these are the subject of Sect.9.9. In Sect.9.10, a second GEE approach is described in which the estimating equations for the mean are supplemented with a second set for the variances/covariances. For GLMMs, extra care must be taken with parameter interpretation, and Sect.9.11 discusses this issue, emphasizing how interpretation differs between conditional and marginal models. In PartII of the book, which focused on independent data, Chap.7 was devoted to models for binary data. For dependent data, models binary data are less well developed, and so we do not devote a complete chapter to their description. However, Sect.9.12 introduces the modeling of dependent binary data, and, subsequently, Sects.9.13 and 9.14 describe conditional (mixed) and marginal models for binary data. Section 9.15 considers how nonlinear models, as defined in Sect.6.10, can be extended to the dependent data case. For such models, many applications concentrate on inference for units, and so the introduction of random effects is again suggested. We refer to the resultant class of models as nonlinear mixed models (NLMMs). Section9.16 considers issues related to the parameterization of the nonlinear model. Inference for nonlinear mixed models via likelihood and Bayes approaches is covered in Sects.9.17 and9.18, while GEE is briefly considered in Sect.9.19. The assessment of assumptions for general regression models is described in Sect.9.20, with concluding comments contained in Sect.9.21. Additional references appear in Sect.9.22.

Soybean [Glycinemax (L.) Merr.] maturity determines the growing region of a given soybean variety and is a primary factor in yield and other agronomic traits. The objectives of this research were to identify the quantitative trait loci (QTL) associated with maturity groups (MGs) and determine the genetic control of soybean maturity in each MG. Using data from 16,879 soybean accessions, genome-wide association (GWA) analyses were conducted for each paired MG and across MGs 000 through IX. Genome-wide association analyses were also performed using 184 genotypes (MGs V-IX) with days to flowering (DTF) and maturity (DTM) collected in the field. A total of 58 QTL were identified to be significantly associated with MGs in individual GWAs, which included 12 reported maturity loci and two stem termination genes. Genome-wide associations across MGs 000-IX detected a total of 103 QTL and confirmed 54 QTL identified in the individual GWAs. Of significant loci identified, qMG-5.2 had effects on the highest number (9) of MGs, followed by E2, E3, Dt2, qMG-15.5, E1, qMG-13.1, qMG-7.1, and qMG-16.1, which affected five to seven MGs. A high number of genetic loci (8-25) that affected MGs 0-V were observed. Stem termination genes Dt1 and Dt2 mainly had significant allele variation in MGs II-V. Genome-wide associations for DTF, DTM, and reproductive period (RP) in the diversity panel confirmed 15 QTL, of which seven were observed in MGs V-IX. The results generated can help soybean breeders manipulate the maturity loci for genetic improvement of soybean yield.

Sasang constitution (SC) medicine, a branch of Korean traditional medicine, classifies the individual into one of four constitutional types (Taeum, TE; Soeum, SE; Soyang, SY; and Taeyang, TY) based on physiologic characteristics. The authors of the current article recently reported individual genetic elements associated with SC types via genome-wide association (GWA) analysis. However, to understand the biologic mechanisms underlying constitution, a comprehensive approach that combines individual genetic effects was applied. Genotypes of 1222 subjects of defined constitution types were measured for 341,998 genetic loci across the entire genome. The biologic pathways associated with SC types were identified via GWA analysis using three different algorithms--namely, the Z-static method, a restandardized gene set assay, and a gene set enrichment assay. Distinct pathways were associated (p<0.05) with each constitution type. The TE type was significantly associated with cytoskeleton-related pathways. The SE type was significantly associated with cardio- and amino-acid metabolism-related pathways. The SY type was associated with enriched melanoma-related pathways. TY subjects were excluded because of the small size of that sample. Among these functionally related pathways, core-node genes regulating multiple pathways were identified. TJP1, PTK2, and SRC were selected as core-nodes for TE; RHOA, and MAOA/MAOB for SE; and GNAO1 for SY (p<0.05), respectively. The current authors systematically identified the biologic pathways and core-node genes associated with SC types from the GWA study; this information should provide insights regarding the molecular mechanisms inherent in constitutional pathophysiology.

Hybrid Bayes factors for genome-wide association studies when a robust test is used

BackgroundLongitudinal ordinal data are commonly analyzed using a marginal proportional odds model for relating ordinal outcomes to covariates in the biomedical and health sciences. The generalized estimating equation (GEE) consistently estimates the regression parameters of marginal models even if the working covariance structure is misspecified. For small-sample longitudinal binary data, recent studies have shown that the bias of regression parameters may result from the GEE and have addressed the issue by applying Firth’s adjustment for the likelihood score equation to the GEE as if generalized estimating functions were likelihood score functions. In this manuscript, for the proportional odds model for longitudinal ordinal data, the small-sample properties of the GEE were investigated, and a bias-reduced GEE (BR-GEE) was derived.MethodsBy applying the adjusted function originally derived for the likelihood score function of the proportional odds model to the GEE, we produced the BR-GEE. We investigated the small-sample properties of both GEE and BR-GEE through simulation and applied them to a clinical study dataset.ResultsIn simulation studies, the BR-GEE had a bias closer to zero, smaller root mean square error than the GEE with coverage probability of confidence interval near or above the nominal level. The simulation also showed that BR-GEE maintained a type I error rate near or below the nominal level.ConclusionsFor the analysis of longitudinal ordinal data involving a small number of subjects, the BR-GEE is advantageous for obtaining estimates of the regression parameters of marginal proportional odds models.

Obesity is a classical complex trait, influenced by both genetic and lifestyle factors. The number of obesity gene variants is currently unknown but, based on sound evolutionary principles, likely to be many, each with a modest effect on the phenotype. Recent advances in our knowledge of variation in the human genome and high throughput genotyping technologies have made possible genome-wide association (GWA) analysis and the identification of bona fide susceptibility genes for many complex diseases and phenotypes, including obesity and its comorbid conditions. GWA analysis in even larger numbers of individuals through collaborative efforts of many investigators will likely identify those polygenes of moderate and modest effect size that manifest in our typical environment. Once the subset of real-world-relevant obesity susceptibility variants is identified, follow-up studies, including detailed molecular analysis of the loci, stratified analyses, prospective and interventional studies in humans, and mechanistic studies in cells and animals will allow us to define the genetic architecture of the locus and dissect how these genes interact with specific environmental and other factors. The molecular and analytical tools to accomplish these goals are now in hand, but cooperation among investigators will be necessary to amass the requisite numbers of phenotyped and genotyped individuals. Identification of susceptibility genes for obesity and determining how they interact with each other and the environment will lead to new insights into the molecular, cellular, and physiological basis of energy homeostasis, and novel strategies for prevention and treatment.

An eggplant (Solanum melongena) association panel of 191 accessions, comprising a mixture of breeding lines, old varieties and landrace selections was SNP genotyped and phenotyped for key breeding fruit and plant traits at two locations over two seasons. A genome-wide association (GWA) analysis was performed using the mixed linear model, which takes into account both a kinship matrix and the sub-population membership of the accessions. Overall, 194 phenotype/genotype associations were uncovered, relating to 30 of the 33 measured traits. These associations involved 79 SNP loci mapping to 39 distinct chromosomal regions distributed over all 12 eggplant chromosomes. A comparison of the map positions of these SNPs with those of loci derived from conventional linkage mapping showed that GWA analysis both validated many of the known controlling loci and detected a large number of new marker/trait associations. Exploiting established syntenic relationships between eggplant chromosomes and those of tomato and pepper recognized orthologous regions in ten eggplant chromosomes harbouring genes influencing breeders’ traits.

In this study, we used the Illumina OvineSNP50 BeadChip to conduct a genome-wide association (GWA) analysis for milk production traits in dairy sheep by analyzing a commercial population of Spanish Churra sheep. The studied population consisted of a total of 1,681 Churra ewes belonging to 16 half-sib families with available records for milk yield (MY), milk protein and fat yields (PY and FY) and milk protein and fat contents (PP and FP). The most significant association identified reached experiment-wise significance for PP and FP and was located on chromosome 3 (OAR3). These results confirm the population-level segregation of a previously reported QTL affecting PP and suggest that this QTL has a significant pleiotropic effect on FP. Further associations were detected at the chromosome-wise significance level on 14 other chromosomal regions. The marker on OAR3 showing the highest significant association was located at the third intron of the alpha-lactalbumin (LALBA) gene, which is a functional and positional candidate underlying this association. Sequencing this gene in the 16 Churra rams of the studied resource population identified additional polymorphisms. One out of the 31 polymorphisms identified was located within the coding gene sequence (LALBA_g.242T>C) and was predicted to cause an amino acid change in the protein (Val27Ala). Different approaches, including GWA analysis, a combined linkage and linkage disequilibrium study and a concordance test with the QTL segregating status of the sires, were utilized to assess the role of this mutation as a putative QTN for the genetic effects detected on OAR3. Our results strongly support the polymorphism LALBA_g.242T>C as the most likely causal mutation of the studied OAR3 QTL affecting PP and FP, although we cannot rule out the possibility that this SNP is in perfect linkage disequilibrium with the true causal polymorphism.