Research Projects

Graph Neural Networks in Biology: A Case Study on MIC
Zhiqian Chen, PhD
Department of Computer Science and Engineerng
Antimicrobial resistance (AMR) poses a significant challenge in healthcare and public health, with organisms such as nontyphoidal Salmonella leading the way due to their escalating resistance to antimicrobial agents. This situation severely complicates the management and containment of diseases, highlighting the urgent need for more effective techniques to assess antimicrobial susceptibility. Conventional methods, including the broth microdilution technique for determining Minimum Inhibitory Concentrations (MICs), are time-consuming and require extensive manual effort. The advent of machine learning (ML) technologies offers a revolutionary approach to predicting MICs, thereby potentially increasing the efficacy of antimicrobial therapies. This paper explores the latest advancements in ML for MIC prediction, focusing on an innovative approach using Graph Neural Networks (GNNs), which could provide a novel insight into the correlation between gene fragment similarities and MIC values. Within this paper, we introduce the K-mer GNN, a novel GNN model designed for MIC prediction. The K-mer GNN model distinctively identifies and incorporates the similarities among k-mers, integrating these insights into GNN alongside k-mer features. This approach not only elevates the precision of MIC predictions but also sheds light on the genomic factors at the k-mer level that drive antimicrobial resistance. Recombinatin and Diversity in Bovine Coronavirus
Florencia Meyer, PhD
Department of Biochemistry, Molecular Biology, Entomology
Bovine coronavirus (BCoV) is an enteropathogenic and respiratory virus often associated with bovine respiratory disease, a multi-factorial costly disease to the cattle industry in the US and globally. At the sequence level, BCoV is closely related to seasonal human coronavirus and coronaviruses in pigs and dogs. BCoV has been isolated from various deer species, which suggests it has zoonotic potential. The goal of our project is to characterize genetic variation in the BCoV genome to gain insights into the temporal and geographic components of its patterns of variation and to assess how recombination contributes to the emergence of new strains. Cross Sectional Study to Determine Risk Factors for Anaplasmosis and Other Endemic Diseases of Cattle
Isaac Jumper, DVM, PhD
Department of Pathobiology and Population Medicine
Cow-calf production is an important sector of the beef industry in Mississippi. Bovine anaplasmosis, bovine leukosis virus, Neospora caninum, bluetongue virus, bovine viral diarrhea virus, and leptospirosis are all production-limiting diseases that affect cow-calf herds in Mississippi. Little information is available describing the within and between herd prevalence of these diseases in Mississippi. Additionally, the role of management, vector-control, and biosecurity factors in occurrence of disease on MS cow-calf operation is not well understood. The objective of this study is to describe the within and between-herd seroprevalence of several production-limiting diseases in MS beef cow-calf herds, as well as identify management, vector-control, and biosecurity practices that are associated with disease in these herds. Assessment of Landscape Disturbance and Climate Factors in Mapping Disease Transmission Risk near Open Cattle Feedlots
Vitor Martins, PhD
Department of Agricultural and Biological Engineering
This research aims to understand and map the risk of mosquito-borne disease transmission around open cattle feedlots in the United States, particularly in states with high concentrations of feedlots, such as North Carolina and Texas. By leveraging geospatial modeling and a range of satellite-derived climate and landscape datasets, the study intends to identify environmental factors affecting mosquito proliferation and potential disease transmission. Key objectives include building baseline information on feedlot locations, integrating climate and landscape variables from satellite data, and developing geospatial models to map disease transmission hotspots. The project will utilize various datasets including precipitation, temperature, humidity, soil moisture, and land cover, among others. Ultimately, the research aims to provide valuable insights into disease transmission risks and support targeted prevention and control measures to sustain cattle production in the US. Seeing Through the Murky Waters: Understanding Catfish Disease Susceptibility as a Function of Behavior and Pond Environmental Conditions
Melanie Boudreau, PhD
Department of Wildlife, Fisheries, and Aquaculture
Commercial production of ictalurid catfishes is the largest aquaculture industry in the United States. During potentially hypoxic pond conditions at night, or during high summer temperatures, catfish may be physiologically more susceptible to disease, although the ability of catfish to behaviorally mitigate this risk is unknown. Our team aims to use biologging technology to further understand the impact of environmental conditions on catfish movement and how that links to disease susceptibility. Optimizing biologger attachment and retention in channel catfish (Ictalurus punctatus)
Melanie Boudreau, PhD
Department of Wildlife, Fisheries, and Aquaculture
Within aquaculture, acceleration can be used as a quantitative surrogate for metabolism and contribute data for energy models and fish welfare indexes. We aimed to build upon previous work by examining how various attachment types (i.e., internal or external attachment) might impact the acceleration-metabolism relationship. For internal attachments, we examined retention time of accelerometer units of varying shape, given catfish propensity to engulf and expel foreign objects. This work helps lay the foundation for various applications in energy models and operational welfare for catfish, providing tools for improving management of aquaculture operations. Optimizing biologgers in free-ranging cattle: implications for inference and cattle management in large grazing systems
Melanie Boudreau, PhD
Department of Wildlife, Fisheries, and Aquaculture
Precision livestock farming is open to using biologging technology to remotely monitor livestock health and behavior. Yet, extensive beef cattle systems appear to be slow to adopt many of these sensors despite the potential benefits in free-ranging herds, potentially due to cost or technical or processing challenges. Our team is working to find the optimum sensor combination and data timeframe that allows for an examination of livestock behavior without impacting inferences gained. We will then apply this knowledge to long-term data that allows us to examine the impact of various cattle management regimes on grazing intensity in large grazing systems. Use of Arthropods Vectors to Classify Cattle Herds by Anaplasmosis Infectious Status
David Smith, DVM., PhD
College of Veterinary Medicine
Bovine anaplasmosis is a blood borne infection of cattle. Infection is common in the southeastern and northwestern United States and emerging into the central region of the US. Long-term feeding of antimicrobial drugs is a common practice for the control of anaplasmosis in endemic regions and is therefore an important concern for antimicrobial stewardship. Currently, epidemiological research to test temporal and spatial factors associated with the spread of anaplasmosis requires costly handling of cattle to collect blood samples to determine herd infection status. The objective of this research is to test the diagnostic value of collecting mechanical (tabanid flies) and biological (Dermacentor ticks) vectors to more easily determine cattle herd infection status with less expense and less risk for injury to cattle. This research is part of the PhD graduate program of Keegan Jones, DVM. Test Performance of Normal Saline as a Transport Medium for Detection of Tritrichomonas foetus in Cattle Herds
David Smith, DVM., PhD
College of Veterinary Mediciine Bovine trichomoniasis is a venereal infection of cattle responsible for significant reproductive losses in infected herds, and an important regulatory disease of cattle in the United States. A recent PCR-based diagnostic test has promise for more accurate classification of infected cattle; however, the test was validated using phosphate-buffered saline (PBS) as the transport media. Even though PBS is a common laboratory reagent, it is not common in veterinary practices and is more expensive than the readily available 0.9% (normal) saline. The objective of this research is to compare the diagnostic performance of PCR with normal saline compared to PBS as the transport medium. This research is part of the MS graduate program of Tyler Jumper, DVM. Using Computer Vision and Radar to Understand and Predict Parasite Spread
Garrett Street, PhD
Department of Wildlife, Fisheries, and Aquaculture
This project focuses on the movements of commercial honey bees as it affects colony health and persistence through the spread of parasitic Varroa mites and the diseases they carry. First, using an AI-driven computer vision system we will monitor marked honey bees within and between colonies to identify the frequency and determinants of movement behaviors contributing to Varroa spread (i.e. drifting, when bees from one colony migrate into another; and robbing, when healthy colonies invade weaker colonies to steal royal jelly and honey). Second, using a novel scanning harmonic radar system, we will monitor the movements of individually tagged bees throughout the landscape to characterize bee movements based on habitat preferences and landscape conditions, and identify how foraging behaviors and movement combine to affect pollination services, the likelihood of encountering pesticides, and overall colony health. Rapid Testing for European Foul Brood in Honey Bees
Kimberly Woodruff, PhD
Departmetn of Clinical Sciences
Testing for American Foul Brood and European Foul Brood require samples from inside the hive which is invasive to the honeybees in the hive. We are proposing to develop a rapid screening using samples that can be collected from outside the hive, for instance, from the outside of the entrance into the hive. Once a less invasive means of disease detection is developed, we can monitor the disease status in multiple areas and map the prevalence and spread of the disease and look for associations of disease and environmental factors. Minimizing Disease Transmission in Poultry through Rapid Detection and Predictive Models
Li Zhang, PhD
Department of Poultry Sciences
This project focuses on advancing poultry health by minimizing disease transmission, with a particular emphasis on addressing clostridial diseases, mycoplasmosis, and avian colibacillosis. First, we'll analyze the genomic characteristics of avian bacterial pathogens by sequencing, aiming to identify specific biomarkers. Second, using these biomarkers, we will develop a rapid, field-deployable detection system based on the loop-mediated isothermal amplification (LAMP) assay for quick disease monitoring. Lastly, we will create predictive models that integrate data from various sources, including climate and livestock density, to forecast infectious diseases in poultry, enhancing biosecurity and management practices against emerging threats.