Translational Bioinformatics for Disease Diagnosis and Treatment Planning
Research Abstract
Microarray technology offers a significant tool for harnessing the true potential of human genome discovery via development of efficient disease diagnosis and treatment planning tools. It measures the expression (activity) levels of genes in living cells, and these levels can be used to characterize cell malfunctions that dictate the manifestations of various diseases. However, noise inherent in the microarray image, which distorts the gene expression values, has been a major obstacle. Research literature presents some approaches for noise reduction. Limited success of these approaches can be attributed partly to the simplicity of the statistical apparatus employed, and the failure to recognize the distinctive characteristics of microarray images. The high cost and unreliability of gene expression values from microarrays have discouraged the development of gene based patterns for disease characterization and treatment planning. Through our research, we intend to improve reliability of gene expression values which will help in achieving the widely anticipated advances in healthcare.
Our research has the following objectives:
- Develop a mathematical framework and an associated experimental and computational methodology to reducing noise level in microarrays. The mathematical framework would be based on a currently evolving theory of Dual Tree Complex Wavelet Transform (DT-CWT).
- Demonstrate the potential of reliable gene expression values in disease diagnosis and treatment strategy planning for prostate cancer, a leading cause of illness and death among men. A gene expression pattern based disease characterization will be far superior to the current approach of using routinely observed clinical parameters. Therefore, develop a methodology for establishing patterns of gene expression that can be used to classify prostate cancer patients and determine corresponding treatment needs.
Publications
Conference Presentations
- “A Novel Approach for Denoising Gene Expression Values Obtained from Microarrays", Presentation made in the Quality, Statistics, and Reliability (QSR) section at an Invited Session at INFORMS annual meeting at Pittsburgh , PA , November 2006.
- "A Multiresolution Approach to Microarray Denoising for Efficient Disease Diagnosis and Drug Discovery”, Poster Presentation made during “Engineering Research Day” at University of South Florida , Tampa , Spring 2007.
|
