Using a Machine Learning Approach to Derive Single Nucleotide Polymorphisms and Microsatellite Cofactors of Cervical Cancer
| Autor: | Kai-Chih Hu, 胡凱智 |
|---|---|
| Rok vydání: | 2003 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 91 Cervical cancer is a common cancer among women worldwide. Recently, infection with high-risk types of human papillomavirus (HPV) has been identified as the central cause of cervical cancer. Although HPV has been identified as playing the central role in cervical cancer, it is still insufficient to confer the cervical cancer. Obviously, there are other environmental and host factors which are involved in the progression of HPV infection to cancer. Recent population based twins and family studies have showed the hereditary component of cervical cancer, indicating genetic susceptibility plays an important role. Thus, to take some SNP markers and microsatellite into account as the genetic factors might be helpful to find out some combinations of these genetic factors and HPV that are involved in the progress of precancerous changes into cervical cancer. We take patients’ age, 11 SNP markers and 4 microsatellites into account and perform decision tree analysis using different learning algorithms. We also compare the performance among these learning algorithms. We anticipate that the results of this study will open the door for investigations of identifying the combinations of genetic factors such as SNPs and microsatellites that interact in a non-additive or nonlinear manner to influence risk of common complex multifactorial disease. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|