Computational & Applied Mathematics & Statistics

Lauren McCluskey, "Matching on What Matters: Improving VCG Selection with Quantifiable Biological Covariates"

Ian Lee, "Divisive Normalization in the Honeybee Bimodal Antennal Lobe"

Session I

Lauren McCluskey, "Matching on What Matters: Improving VCG Selection with Quantifiable Biological Covariates"

Nonclinical studies traditionally require concurrent control animals comprising approximately 25% of the total study population, contributing to both cost and ethical concerns. Virtual Control Groups (VCGs) offer a promising alternative by substituting historical control data for concurrent controls. However, selecting appropriate VCGs is challenging, as they must match the recipient study as closely as possible, and established thresholds for acceptable variability in parameters such as body weight and age are not well-defined. This project produced a Spotfire dashboard to aid Toxicologists, Pathologists, and Study Directors in selecting viable VCGs based on quantifiable biological covariates. Nonclinical Study Data is analyzed and filtered to select a set of VCGs that can be integrated into the recipient study. This system will continue to improve over time as additional studies are conducted and entered into the data warehouse.

Student Major(s): Computational & Applied Mathematics & Statistics (Mathematical Biology Track) and Biology Major

Advisor: Wenxian Wang

Video Presentations

Ian Lee, "Divisive Normalization in the Honeybee Bimodal Antennal Lobe"

The antennal lobe of the honeybee is a part of their brain that is necessary for them to track food sources and other important signals. A challenge they face is that the sensory environment is dynamic and unpredictable, meaning their inputs need to be appropriately scaled to fire at appropriate levels through vastly varying amounts of input. One proposed mechanism of this is through divisive normalization, which scales the strength of the input on how active neighboring neurons are, to strengthen inputs when at low stimulation environments, and weaken inputs when in noisy environments. This was tested using an integrate and fire model of the antennal lobe based on previous research, which is made up of glomeruli containing projection neurons and local neurons, and receives both odor and mechanosensory input. We found that divisive normalization occurs in this model of the antennal lobe.

Student Major(s)/Minor: Neuroscience Major

Advisor: Dr. Mainak Patel

Page updated

Google Sites

Report abuse