BG2: Bayesian variable selection in generalized linear mixed models with nonlocal priors for non-Gaussian GWAS data.

BACKGROUND: Genome-wide association studies (GWASes) aim to identify single nucleotide polymorphisms (SNPs) associated with a given phenotype. A common approach for the analysis of GWAS is single marker analysis (SMA) based on linear mixed models (LMMs). However, LMM-based SMA usually yields a large number of false discoveries and cannot be directly applied to non-Gaussian phenotypes such as count data. RESULTS: We present a novel Bayesian method to find SNPs associated with non-Gaussian phenotypes. To that end, we use generalized linear mixed models (GLMMs) and, thus, call our method Bayesian GLMMs for GWAS (BG2). To deal with the high dimensionality of GWAS analysis, we propose novel nonlocal priors specifically tailored for GLMMs. In addition, we develop related fast approximate Bayesian computations. BG2 uses a two-step procedure: first, BG2 screens for candidate SNPs; second, BG2 performs model selection that considers all screened candidate SNPs as possible regressors. A simulation study shows favorable performance of BG2 when compared to GLMM-based SMA. We illustrate the usefulness and flexibility of BG2 with three case studies on cocaine dependence (binary data), alcohol consumption (count data), and number of root-like structures in a model plant (count data).

Bayesian model selection for generalized linear mixed models.

We propose a Bayesian model selection approach for generalized linear mixed models (GLMMs). We consider covariance structures for the random effects that are widely used in areas such as longitudinal studies, genome-wide association studies, and spatial statistics. Since the random effects cannot be integrated out of GLMMs analytically, we approximate the integrated likelihood function using a pseudo-likelihood approach. Our Bayesian approach assumes a flat prior for the fixed effects and includes both approximate reference prior and half-Cauchy prior choices for the variances of random effects. Since the flat prior on the fixed effects is improper, we develop a fractional Bayes factor approach to obtain posterior probabilities of the several competing models. Simulation studies with Poisson GLMMs with spatial random effects and overdispersion random effects show that our approach performs favorably when compared to widely used competing Bayesian methods including deviance information criterion and Watanabe-Akaike information criterion. We illustrate the usefulness and flexibility of our approach with three case studies including a Poisson longitudinal model, a Poisson spatial model, and a logistic mixed model. Our proposed approach is implemented in the R package GLMMselect that is available on CRAN.

BGWAS: Bayesian variable selection in linear mixed models with nonlocal priors for genome-wide association studies.

BACKGROUND: Genome-wide association studies (GWAS) seek to identify single nucleotide polymorphisms (SNPs) that cause observed phenotypes. However, with highly correlated SNPs, correlated observations, and the number of SNPs being two orders of magnitude larger than the number of observations, GWAS procedures often suffer from high false positive rates. RESULTS: We propose BGWAS, a novel Bayesian variable selection method based on nonlocal priors for linear mixed models specifically tailored for genome-wide association studies. Our proposed method BGWAS uses a novel nonlocal prior for linear mixed models (LMMs). BGWAS has two steps: screening and model selection. The screening step scans through all the SNPs fitting one LMM for each SNP and then uses Bayesian false discovery control to select a set of candidate SNPs. After that, a model selection step searches through the space of LMMs that may have any number of SNPs from the candidate set. A simulation study shows that, when compared to popular GWAS procedures, BGWAS greatly reduces false positives while maintaining the same ability to detect true positive SNPs. We show the utility and flexibility of BGWAS with two case studies: a case study on salt stress in plants, and a case study on alcohol use disorder. CONCLUSIONS: BGWAS maintains and in some cases increases the recall of true SNPs while drastically lowering the number of false positives compared to popular SMA procedures.

Mapping Genetic Variation in Arabidopsis in Response to Plant Growth-Promoting Bacterium Azoarcus olearius DQS-4T.

Plant growth-promoting bacteria (PGPB) can enhance plant health by facilitating nutrient uptake, nitrogen fixation, protection from pathogens, stress tolerance and/or boosting plant productivity. The genetic determinants that drive the plant-bacteria association remain understudied. To identify genetic loci highly correlated with traits responsive to PGPB, we performed a genome-wide association study (GWAS) using an Arabidopsis thaliana population treated with Azoarcus olearius DQS-4T. Phenotypically, the 305 Arabidopsis accessions tested responded differently to bacterial treatment by improving, inhibiting, or not affecting root system or shoot traits. GWA mapping analysis identified several predicted loci associated with primary root length or root fresh weight. Two statistical analyses were performed to narrow down potential gene candidates followed by haplotype block analysis, resulting in the identification of 11 loci associated with the responsiveness of Arabidopsis root fresh weight to bacterial inoculation. Our results showed considerable variation in the ability of plants to respond to inoculation by A. olearius DQS-4T while revealing considerable complexity regarding statistically associated loci with the growth traits measured. This investigation is a promising starting point for sustainable breeding strategies for future cropping practices that may employ beneficial microbes and/or modifications of the root microbiome.

Long term temporal trends in synoptic-scale weather conditions favoring significant tornado occurrence over the central United States.

We perform a statistical climatological study of the synoptic- to meso-scale weather conditions favoring significant tornado occurrence to empirically investigate the existence of long term temporal trends. To identify environments that favor tornadoes, we apply an empirical orthogonal function (EOF) analysis to temperature, relative humidity, and winds from the Modern-Era Retrospective analysis for Research and Applications Version 2 (MERRA-2) dataset. We consider MERRA-2 data and tornado data from 1980 to 2017 over four adjacent study regions that span the Central, Midwestern, and Southeastern United States. To identify which EOFs are related to significant tornado occurrence, we fit two separate groups of logistic regression models. The first group (LEOF models) estimates the probability of occurrence of a significant tornado day (EF2-EF5) within each region. The second group (IEOF models) classifies the intensity of tornadic days either as strong (EF3-EF5) or weak (EF1-EF2). When compared to approaches using proxies such as convective available potential energy, our EOF approach is advantageous for two main reasons: first, the EOF approach allows for the discovery of important synoptic- to mesoscale variables previously not considered in the tornado science literature; second, proxy-based analyses may not capture important aspects of three-dimensional atmospheric conditions represented by the EOFs. Indeed, one of our main novel findings is the importance of a stratospheric forcing mode on occurrence of significant tornadoes. Other important novel findings are the existence of long-term temporal trends in the stratospheric forcing mode, in a dry line mode, and in an ageostrophic circulation mode related to the jet stream configuration. A relative risk analysis also indicates that changes in stratospheric forcings are partially or completely offsetting increased tornado risk associated with the dry line mode, except in the eastern Midwest region where tornado risk is increasing.

Long-term recovery from opioid use disorder: recovery subgroups, transition states and their association with substance use, treatment and quality of life.

BACKGROUND AND AIMS: Limited information exists regarding individual subgroups of recovery from opioid use disorder (OUD) following treatment and how these subgroups may relate to recovery trajectories. We used multi-dimensional criteria to identify OUD recovery subgroups and longitudinal transitions across subgroups. DESIGN, SETTING AND PARTICIPANTS: In a national longitudinal observational study in the United States, individuals who previously participated in a clinical trial for subcutaneous buprenorphine injections for treatment of OUD were enrolled and followed for an average of 4.2 years after participation in the clinical trial. MEASUREMENTS: We identified recovery subgroups based on psychosocial outcomes including depression, opioid withdrawal and pain. We compared opioid use, treatment utilization and quality of life among these subgroups. FINDINGS: Three dimensions of the recovery process were identified: depression, opioid withdrawal and pain. Using these three dimensions, participants were classified into four recovery subgroups: high-functioning (minimal depression, mild withdrawal and no/mild pain), pain/physical health (minimal depression, mild withdrawal and moderate pain), depression (moderate depression, mild withdrawal and mild/moderate pain) and low-functioning (moderate/severe withdrawal, moderate depression and moderate/severe pain). Significant differences among subgroups were observed for DSM-5 criteria (P < 0.001) and remission status (P < 0.001), as well as with opioid use (P < 0.001), treatment utilization (P < 0.001) and quality of life domains (physical health, psychological, environment and social relationships; Ps < 0.001, Cohen's fs ≥ 0.62). Recovery subgroup assignments were dynamic, with individuals transitioning across subgroups during the observational period. Moreover, the initial recovery subgroup assignment was minimally predictive of long-term outcomes. CONCLUSIONS: There appear to be four distinct subgroups among individuals in recovery from OUD. Recovery subgroup assignments are dynamic and predictive of contemporaneous, but not long-term, substance use, substance use treatment utilization or quality of life outcomes.

BICOSS: Bayesian iterative conditional stochastic search for GWAS.

BACKGROUND: Single marker analysis (SMA) with linear mixed models for genome wide association studies has uncovered the contribution of genetic variants to many observed phenotypes. However, SMA has weak false discovery control. In addition, when a few variants have large effect sizes, SMA has low statistical power to detect small and medium effect sizes, leading to low recall of true causal single nucleotide polymorphisms (SNPs). RESULTS: We present the Bayesian Iterative Conditional Stochastic Search (BICOSS) method that controls false discovery rate and increases recall of variants with small and medium effect sizes. BICOSS iterates between a screening step and a Bayesian model selection step. A simulation study shows that, when compared to SMA, BICOSS dramatically reduces false discovery rate and allows for smaller effect sizes to be discovered. Finally, two real world applications show the utility and flexibility of BICOSS. CONCLUSIONS: When compared to widely used SMA, BICOSS provides higher recall of true SNPs while dramatically reducing false discovery rate.

Musashi1 Contribution to Glioblastoma Development via Regulation of a Network of DNA Replication, Cell Cycle and Division Genes.

RNA-binding proteins (RBPs) function as master regulators of gene expression. Alterations in their levels are often observed in tumors with numerous oncogenic RBPs identified in recent years. Musashi1 (Msi1) is an RBP and stem cell gene that controls the balance between self-renewal and differentiation. High Msi1 levels have been observed in multiple tumors including glioblastoma and are often associated with poor patient outcomes and tumor growth. A comprehensive genomic analysis identified a network of cell cycle/division and DNA replication genes and established these processes as Msi1's core regulatory functions in glioblastoma. Msi1 controls this gene network via two mechanisms: direct interaction and indirect regulation mediated by the transcription factors E2F2 and E2F8. Moreover, glioblastoma lines with Msi1 knockout (KO) displayed increased sensitivity to cell cycle and DNA replication inhibitors. Our results suggest that a drug combination strategy (Msi1 + cell cycle/DNA replication inhibitors) could be a viable route to treat glioblastoma.

Modeling allele-specific expression at the gene and SNP levels simultaneously by a Bayesian logistic mixed regression model.

BACKGROUND: High-throughput sequencing experiments, which can determine allele origins, have been used to assess genome-wide allele-specific expression. Despite the amount of data generated from high-throughput experiments, statistical methods are often too simplistic to understand the complexity of gene expression. Specifically, existing methods do not test allele-specific expression (ASE) of a gene as a whole and variation in ASE within a gene across exons separately and simultaneously. RESULTS: We propose a generalized linear mixed model to close these gaps, incorporating variations due to genes, single nucleotide polymorphisms (SNPs), and biological replicates. To improve reliability of statistical inferences, we assign priors on each effect in the model so that information is shared across genes in the entire genome. We utilize Bayesian model selection to test the hypothesis of ASE for each gene and variations across SNPs within a gene. We apply our method to four tissue types in a bovine study to de novo detect ASE genes in the bovine genome, and uncover intriguing predictions of regulatory ASEs across gene exons and across tissue types. We compared our method to competing approaches through simulation studies that mimicked the real datasets. The R package, BLMRM, that implements our proposed algorithm, is publicly available for download at https://github.com/JingXieMIZZOU/BLMRM . CONCLUSIONS: We will show that the proposed method exhibits improved control of the false discovery rate and improved power over existing methods when SNP variation and biological variation are present. Besides, our method also maintains low computational requirements that allows for whole genome analysis.

miR-124, -128, and -137 Orchestrate Neural Differentiation by Acting on Overlapping Gene Sets Containing a Highly Connected Transcription Factor Network.

The ventricular-subventricular zone harbors neural stem cells (NSCs) that can differentiate into neurons, astrocytes, and oligodendrocytes. This process requires loss of stem cell properties and gain of characteristics associated with differentiated cells. miRNAs function as important drivers of this transition; miR-124, -128, and -137 are among the most relevant ones and have been shown to share commonalities and act as proneurogenic regulators. We conducted biological and genomic analyses to dissect their target repertoire during neurogenesis and tested the hypothesis that they act cooperatively to promote differentiation. To map their target genes, we transfected NSCs with antagomiRs and analyzed differences in their mRNA profile throughout differentiation with respect to controls. This strategy led to the identification of 910 targets for miR-124, 216 for miR-128, and 652 for miR-137. The target sets show extensive overlap. Inspection by gene ontology and network analysis indicated that transcription factors are a major component of these miRNAs target sets. Moreover, several of these transcription factors form a highly interconnected network. Sp1 was determined to be the main node of this network and was further investigated. Our data suggest that miR-124, -128, and -137 act synergistically to regulate Sp1 expression. Sp1 levels are dramatically reduced as cells differentiate and silencing of its expression reduced neuronal production and affected NSC viability and proliferation. In summary, our results show that miRNAs can act cooperatively and synergistically to regulate complex biological processes like neurogenesis and that transcription factors are heavily targeted to branch out their regulatory effect.

Bayesian hierarchical multi-subject multiscale analysis of functional MRI data.

We develop a methodology for Bayesian hierarchical multi-subject multiscale analysis of functional Magnetic Resonance Imaging (fMRI) data. We begin by modeling the brain images temporally with a standard general linear model. After that, we transform the resulting estimated standardized regression coefficient maps through a discrete wavelet transformation to obtain a sparse representation in the wavelet space. Subsequently, we assign to the wavelet coefficients a prior that is a mixture of a point mass at zero and a Gaussian white noise. In this mixture prior for the wavelet coefficients, the mixture probabilities are related to the pattern of brain activity across different resolutions. To incorporate this information, we assume that the mixture probabilities for wavelet coefficients at the same location and level are common across subjects. Furthermore, we assign for the mixture probabilities a prior that depends on a few hyperparameters. We develop an empirical Bayes methodology to estimate the hyperparameters and, as these hyperparameters are shared by all subjects, we obtain precise estimated values. Then we carry out inference in the wavelet space and obtain smoothed images of the regression coefficients by applying the inverse wavelet transform to the posterior means of the wavelet coefficients. An application to computer simulated synthetic data has shown that, when compared to single-subject analysis, our multi-subject methodology performs better in terms of mean squared error. Finally, we illustrate the utility and flexibility of our multi-subject methodology with an application to an event-related fMRI dataset generated by Postle (2005) through a multi-subject fMRI study of working memory related brain activation.

Construçäo de uma matriz de contabilidade social para o Brasil / Construction of a social accounting to th Brazil

Relata que a matriz de contabilidade social (MCS) é uma representação estilizada da totalidde dos fluxos de recursos de uma economia em um certo ano. Explica que tem por finalidade informar sobre um modelo de análise da inter-relação entre crescimento econômico, composição do produto e do emprego e desigualdade da distribuição da renda. Além da construção da MCS, apresenta metodologias para a estimativa da composição dos investimentos setoriais a partir de tabelas auxiliares da matriz insumo produto (MIP) e outros resultados complementares. Descreve a MCS proposta. Comenta os problemas e as soluções adotadas na construção desta matriz e os procedimentos adotados nesta construção.