TITLE: “Baseline MEthylation Patterns prior to flu vaccination.”
ABSTRACT: Influenza affects millions worldwide each year with responses varying from individual to individual. Influenza can be broken down into subtypes specifically H1N1, H3N2, Yamagata, and Victoria. One way to measure the immune response to influenza is to measure a person’s antibody response to influenza. To measure how many antibodies are present in a sample, a hemagglutination inhibition assay (HAI) is used. DNA methylation is an epigenetic mechanism used to regulate gene expression in cells. Its mechanism of action is the addition of a methyl group to cytosine at a cytosine-guanine pair. DNA methylation has been shown to change in response to stimuli such as viral or bacterial infections. DNA methylation can be measured by bisulfite sequencing, specifically reduced representation bisulfite sequencing in our case. In this study, data was taken from patients who had the flu vaccination. Their antibody data was measured using the HAI assay by the University of Georgia and their methylation data was measured using reduced representation bisulfite sequencing by the Pellegrini and Reed lab at UCLA. Using various statistical learning algorithms we were able to find methylated sites that were good predictors of vaccine response. Elastic net regression proved to be a particularly good predictor of vaccine response, and after further analysis, it was revealed that the best prediction happened with only a few significant sites. Some of these significant sites seem to be involved in regulating immune response and membrane function. Further work will be done to determine the prediction accuracy of these algorithms with just these sites. Ideally, after this future work and other experiments, these methylated sites can be used as biomarkers to indicate response to flu vaccination.
