Bioinformatics Seminar
Time: 11AM
Venue: Davis Auditorium and Online
19 November 2024
This is a WEHI only event.Deep learning of the B regulome empowered by edgeR
Junli NieWEHI Bioinformatics/Immunology
B cells afford protective immunity by producing antibodies against pathogens after infection or vaccination. B cells derive from haematopoietic stem cells and ultimately become antibody-secreting cells through multiple stages. Abnormalities at any B cell stage can lead to severe immunodeficiency, blood cancers or autoimmune diseases. Although many of these diseases have a genetic basis, most of the disease-associated mutations are in non-coding regions, which lack functional annotation to date, making it challenging to dissect disease mechanisms. Training deep learning models to predict the functions of mutations in non-coding regions will potentially be a game-changing advance for identifying causal mutations. In this presentation, I will show how I use edgeR to perform time-course analysis and allele-specific analysis for B cell non-coding regions and how I use the edgeR analysis output to train and test CNN models which are biologically explainable and are able to predict the function of B cell non-coding regions.