|
M.S. Defense Thursday, May 31, 2007 Room 216 ENG II 2:30PM Title: Gene Set Clustering Analysis with Adaptive Learning Abstract: Today, DNA microarray analysis is a routine tool in biomedical research, resulting in significant increases in acquired experimental data. Yet, the interpretation of experimental results, involving the translation of data into useful biological knowledge, still remains a major challenge. As different experiments yield varying outcomes, even experiments performed on a common disease, the challenge of collecting and disseminating resulting knowledge becomes more difficult. In the current work, a new system called Gene Set Clustering Analysis (GSCA), facilitates the analysis of expression profiles based on predefined sets of genes sharing common biological functions. In contrast to existing single gene analyses, the new method allows for the interpretation of biological results through a unified biological theme. And, unlike similar modern approaches, GSCA utilizes the adaptive learning model, storing previous knowledge and user feedback in an experience database—a centralized repository for information—which can then be queried in future analyses. The effectiveness of GSCA has been confirmed by analyzing multiple cancer related datasets including breast cancer, lung cancer, and leukemia and comparing results against current literature and established scientific knowledge. |
| < Prev | Next > |
|---|