Develop the Disease Specific Bioinformatics Platforms with Integrated Bioinformatics Data

Abstract

With the advance of multiple types of omics technology and corresponding analytical methods, various type of bioinformatic data have become available. Mining and integrating these data for analysis will provide valuable insights for disease mechanism investigation, drug target identification and new drug development. However, most of the omics data are large size, heterogeneous, and complex, it is challenging for biomedical researchers to mine the data for relevant evidence, especially for those with limited computational skills. In this thesis, I aimed to develop disease specific platforms integrated with multimodal bioinformatic data types to provide researchers with strong bioinformatics support. To achieve this goal, I explored advanced transcriptomic data analytical methods and proposed a novel biomarker for the prediction of overall survival of colon cancer patients, then prototyped a user-friendly patient oriented clinical decision support system to provide accurate and intuitive colorectal cancer risk factor assessment. With the experience of the transcriptomic data analytical methods and the web-based application development, I further designed and implemented Cancer Gene and Pathway Explorer which is an integrative bioinformatics webserver that can be used for cancer publication trends investigation, gene set enrichment analysis with integrated data, and optimal cancer cell line identification. Based on the framework of CGPE, I developed another bioinformatics platform focusing on Alzheimer’s disease, called Alzheimer’s Disease Explorer, which is a first-of-its-kind bioinformatics server, providing rich bioinformatic support from literature, omics and chemical data to facilitate researchers in ND drug development field. By accomplishing a series of work in my thesis, I have shown that integrated disease specific bioinformatics platforms can provide great value to the research community by allowing 1.) fast and accurate investigation of currently available literature, 2.) quick hypothesis generation and validation using transcriptomic datasets, 3.) multi-dimension drug target evaluation and 4) fast querying of published bioinformatics outcomes.