Life Sciences Commons^™

Bayesian Methods For Graphical Models With Neighborhood Selection., Sagnik Bhadury

Electronic Theses and Dissertations

Graphical models determine associations between variables through the notion of conditional independence. Gaussian graphical models are a widely used class of such models, where the relationships are formalized by non-null entries of the precision matrix. However, in high-dimensional cases, covariance estimates are typically unstable. Moreover, it is natural to expect only a few significant associations to be present in many realistic applications. This necessitates the injection of sparsity techniques into the estimation method. Classical frequentist methods, like GLASSO, use penalization techniques for this purpose. Fully Bayesian methods, on the contrary, are slow because they require iteratively sampling over a quadratic …

Computing For Numeracy: How Safe Is Your Covid-19 Social Bubble?, Charles Connor

Numeracy

The COVID-19 pandemic has led many people to form social bubbles. These social bubbles are small groups of people who interact with one another but restrict interactions with the outside world. The assumption in forming social bubbles is that risk of infection and severe outcomes, like hospitalization, are reduced. How effective are social bubbles? A Bayesian event tree is developed to calculate the probabilities of specific outcomes, like hospitalization, using example rates of infection in the greater community and example prior functions describing the effectiveness of isolation by members of the social bubble. The probabilities are solved for two contrasting …

Modified-Half-Normal Distribution And Different Methods To Estimate Average Treatment Effect., Jingchao Sun

Electronic Theses and Dissertations

This dissertation consists of three projects related to Modified-Half-Normal distribution and causal inference. In my first project, a new distribution called Modified-Half-Normal distribution was introduced. I explored a few of its distributional properties, the procedures for generating random samples based on Bayesian approaches, and the parameter estimation based on the method of moments. The second project deals with the problem of selection bias of average treatment effect (ATE) if we use the observational data. I combined the propensity score based inverse probability of treatment weighting (IPTW) method and the directed acyclic graph (DAG) to solve this problem. The third project …

Habitat Associations And Reproduction Of Fishes On The Northwestern Gulf Of Mexico Shelf Edge, Elizabeth Marie Keller

LSU Doctoral Dissertations

Several of the northwestern Gulf of Mexico (GOM) shelf-edge banks provide critical hard bottom habitat for coral and fish communities, supporting a wide diversity of ecologically and economically important species. These sites may be fish aggregation and spawning sites and provide important habitat for fish growth and reproduction. Already designated as habitat areas of particular concern, many of these banks are also under consideration for inclusion in the expansion of the Flower Garden Banks National Marine Sanctuary. This project aimed to gain a more comprehensive understanding of the communities and fish species on shelf-edge banks by way of gonad histology, …

Uncovering The Drivers Of Non-Native Plant Invasions Using Ecological Data Synthesis, Marina Golivets

Graduate College Dissertations and Theses

Understanding what promotes invasiveness of species outside their native range and predicting which ecosystems and under which conditions will be invaded is an ultimate goal of the field of invasion ecology. Obtaining general answers to these questions requires synthesis of extensive yet heterogeneous empirical evidence, coupled with a solid theoretical background. In this dissertation, I sought to provide insight into the drivers of non-native plant invasions through combining and synthesizing ecological data from various sources using advanced statistical techniques. The results of this work are presented as three independent research studies.

In the first study, I aimed to understand what …

Multivariate Analysis Of The Cotton Seed Ionome Reveals A Shared Genetic Architecture, Duke Pauli, Greg Ziegler, Min Ren, Matthew A. Jenks, Douglas J. Hunsaker, Min Zhang, Ivan Baxter, Michael A. Gore

Faculty & Staff Scholarship

To mitigate the effects of heat and drought stress, a better understanding of the genetic control of physiological responses to these environmental conditions is needed. To this end, we evaluated an upland cotton (Gossypium hirsutum L.) mapping population under water-limited and well-watered con- ditions in a hot, arid environment. The elemental concentrations (ionome) of seed samples from the pop- ulation were profiled in addition to those of soil samples taken from throughout the field site to better model environmental variation. The elements profiled in seeds exhibited moderate to high heritabilities, as well as strong phenotypic and genotypic correlations between elements …

Modelling Bird Migration With Motus Data And Bayesian State-Space Models, Justin Baldwin

Masters Theses

Bird migration is a poorly-known yet important phenomenon, as understanding movement patterns of birds can inform conservation strategies and public health policy for animal-borne diseases. Recent advances in wildlife tracking technology, in particular the Motus system, have allowed researchers to track even small flying birds and insects with radio transmitters that weigh fractions of a gram. This system relies on a community-based distributed sensor network that detects tagged animals as they move through the detection nodes on journeys that range from small local movements to intercontinental migrations. The quantity of data generated by the Motus system is unprecedented, is on …

Methods To Account For Breed Composition In A Bayesian Gwas Method Which Utilizes Haplotype Clusters, Danielle F. Wilson-Wells

Department of Statistics: Dissertations, Theses, and Student Work

In livestock, prediction of an animal’s genetic merit using genomic information is becoming increasingly common. The models used to make these predictions typically assume that we are sampling from a homogeneous population. However, in both commercial and experimental populations the sire and dam of an individual may be a mixture of different breeds. Haplotype models can capture this population structure.

Two models based on breed specific haplotype clusters where developed to account for differences across multiple breeds. The first model utilizes the breed composition of the individual, while the second utilizes the breed composition from the sire and dam. Haplotype …

A Bayesian Gwas Method Utilizing Haplotype Clusters For A Composite Breed Population, Danielle F. Wilson-Wells, Stephen D. Kachman

Conference on Applied Statistics in Agriculture

Commercial beef cattle are often composites of multiple breeds. Current methods used to produce genomic predictors are based on the underlying assumption of animals being sampled from a homogeneous population. As a result, the predictors can perform poorly when used to predict the relative genetic merit of animals whose breed composition are different. In part, this is due to the changes in linkage disequilibrium between the markers and the quantitative trait loci as we move from one breed to the next. An alternative model based on breed specific haplotype clusters was developed to allow for differences in linkage disequilibrium across …

Germline Mutation Detection In Next Generation Sequencing Data And Tp53 Mutation Carrier Probability Estimation For Li-Fraumeni Syndrome, Gang Peng

Dissertations & Theses (Open Access)

Next generation sequencing technology has been widely used in genomic analysis, but its application has been compromised by the missing true variants, especially when these variants are rare. We proposed a family-based variant calling method, FamSeq, integrating Mendelian transmission information with de novo mutation and sequencing data to improve the variant calling accuracy. We investigated the factors impacting the improvement of family-based variant calling in simulation data and validated it in real sequencing data. In both simulation and real data, FamSeq works better than the single individual based method.

In FamSeq, we implemented four different methods for the Mendelian genetic …

Atmospheric Tomography: A Bayesian Inversion Technique For Determining The Rate And Location Of Fugitive Emissions, Ruhi Humphries, Charles Jenkins, Ray Leuning, Steve Zegelin, David Griffith, Christopher Caldow, Henry Berko, Andrew Feitz

Faculty of Science - Papers (Archive)

A Bayesian inversion technique to determine the location and strength of trace gas emissions from a point source in open air is presented. It was tested using atmospheric measurements of N2O and CO2 released at known rates from a source located within an array of eight evenly spaced sampling points on a 20 m radius circle. The analysis requires knowledge of concentration enhancement downwind of the source and the normalized, three-dimensional distribution (shape) of concentration in the dispersion plume. The influence of varying background concentrations of ~1% for N2O and ~10% for CO2 was removed by subtracting upwind concentrations from …

Bi-Level Clustering Of Mixed Categorical And Numerical Biomedical Data, Bill Andreopoulos, Aijun An, Xiaogang Wang

Faculty Publications, Computer Science

Biomedical data sets often have mixed categorical and numerical types, where the former represent semantic information on the objects and the latter represent experimental results. We present the BILCOM algorithm for |Bi-Level Clustering of Mixed categorical and numerical data types|. BILCOM performs a pseudo-Bayesian process, where the prior is categorical clustering. BILCOM partitions biomedical data sets of mixed types, such as hepatitis, thyroid disease and yeast gene expression data with Gene Ontology annotations, more accurately than if using one type alone.

Bi-Level Clustering Of Mixed Categorical And Numerical Biomedical Data, Bill Andreopoulos, Aijun An, Xiaogang Wang

William B. Andreopoulos

Vertical Integration In The Chicken Broiler Industry, Juana Sanchez

Conference on Applied Statistics in Agriculture

This paper analyzes three hypotheses concerning supply in the U.S. chicken broiler industry: (a) there has been a cycle in the industry of approximately 27-36 months length; (b) the seasonal and other periodic components, as well as relations between variables, have changed as a result of vertical integration in the industry; (c) the effects of vertical integration in the industry were counteracted in the early seventies by such forces external to the industry as domestic and international economic conditions .

The hypotheses are analyzed using new monthly, non-seasonally adjusted time series data for chick placement, wholesale broiler prices, chicks hatched …