Committee: 

Dr. M.D. Wang, Advisor        

Dr. Inan, Chair

Dr. QiuButera

Abstract: The objective of the proposed research is to investigate, develop, and validate robust bioinformatics approaches for extracting, discovering, and integrating molecular knowledge using next-generation sequencing data. RNA sequencing, or RNA-seq for short, is a major branch of the next-generation sequencing technology. To carry out this objective, we will first investigate the impact of bioinformatics pipeline components on gene expression estimation and downstream analysis using RNA-seq data. Through such the investigation, we will identify variations in downstream expression estimation and prediction performance induced by pipeline components. To ameliorate such the variations or to remedy poor-performing pipelines, we will implement filtering-based approaches so as to avoid skeptical information propagates through the pipeline as well as to implement a novel normalization method that can mitigate biases and improve the robustness of the RNA-seq expression analysis pipeline. Lastly, we aim to address the curse of dimensionality and to improve the statistical power by developing a novel batch effect correction method so that similar data from multiple sources can be merged as well as developing prediction models that can use multi-modal data as the input. With these novel techniques, we expect to achieve improved biological knowledge discovery and clinical endpoint prediction.