Biostatistics Weekly Seminar

The increasing availability of genomic data offers us opportunities to develop novel statistical and machine-learning methods to enhance our understanding of the molecular basis of human disease and enable effective treatments. However, genomic data pose several challenges for analysis, including excess zeros, noisiness, and spatial relationships. In this talk, I will present two applications that tackle the challenges arising from genomic data, using statistical methods and machine learning approaches, respectively. In the first application, we focus on the modeling of zero-inflated count data with measurement error. We demonstrate the implementation of the method in delineating the association between copy number variants and tumor stages, using multi-institutional genomic data with different data qualities. In the second application, we propose a deep learning algorithm to discover the spatial location of cells in scRNA-seq data. This framework can be applied to study the changes in cell distribution in cerebral cortex layers during the progression of Alzheimer's disease.

Virtual: Zoom Link to Follow
30 August 2023
1:30pm