Date of Graduation

2004

Document Type

Thesis

Degree Type

MS

Committee Chair

Bojan B. Cukic

Abstract

The increasing amount of easily available genomic data has heightened the importance of computational tools in the analysis genome databases, especially the human genome. In this study, we develop a new methodology based on the suffix tree data structure to identify unknown targets of gene families. We use the identification of hypoxia-inducible factor 1 (HIF-1) as a model system. HIF-1 is important in human normal physiological functions and in many human diseases such as cancer and cardiovascular diseases. We selected 17 known HIF-1 target gene sequences as a training set. We use the suffix tree data structure to extract a common set of patterns from the training set. Then the common set of patterns and the 46 known HIF-1 binding sequences are used to search the genome nucleotide database using the suffix tree algorithm. The outputs from the above steps are further processed by position analysis according to gene functional location. Since most known HIF-1 target sequences are at the position of 50 promoter, we selected the potential HIF-1 targets located only at the 50 region of the promoter. We obtained 50 known HIF-1 targets including training genes, and 14,087 potentially novel HIF-1 targets. These results demonstrate that our methodology could be used to identify HIF-1 targets and to predict potentially novel HIF-1 targets. In order to further prove that the potential novel targets are functional HIF-1 targets, we randomly selected one potentially novel HIF-1 to analyze its biological function in the lab. We confirmed the target is a biologically relevant HIF-1 target. From the results of the known HIF-1 genes that are not inside the training set and the new HIF-1 gene we discover, we demonstrate that the methodology is a new computational method for genome search and discovery. By using HIF-1 as a model system to identify novel targets of the transcription factor, we found new HIF-1 target genes that may have important functions for human health and disease research.

Share

COinS