Date of Award


Document Type

Thesis campus only



First Advisor

Luis D Giavedoni

Second Advisor

Kevin Livingstone


RNA-Seq is a high-throughput technique that uses next-generation sequencing of mRNA to quantify expression of genes or gene transcripts. Currently, there is not a clear consensus regarding which data analysis methods are most appropriate for RNA-Seq experiments. Previous studies have shown that different methods, when applied to the same raw data, can produce different conclusions, including different fold-change and false discovery rate (FDR) values. However, these studies did not proceed to the downstream applications of their data, such as Ingenuity Pathway Analysis (IPA) or gene ontology (GO) analysis, that are often the end-goal of many transcriptomics studies to assess the biological relevance of the differentially expressed genes (DEGs). We performed RNA-Seq analyses of mRNA isolated from PBMCs of baboons infected with Zika virus (ZIKV) at days 0, 3, and 15 post-infection, using two currently popular R packages for RNA-seq analyses: edgeR and DESeq2. Within these packages, we compared results obtained from different combinations of statistical methods. We found that these variations in analysis methods led to fairly large changes in the number of DEGs detected, which conforms to the results of previous studies. Interestingly, in most cases, these differences did not appear to drastically change results of Biological Process Gene Ontology analyses performed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) Functional Annotation tool. The results of our GO analyses also indicated significant up- and down- regulation of GO terms related to inflammation and antiviral immune responses, as we expected based on plasma viral load data.