Dissertation > Excellent graduate degree dissertation topics show

Breast cancer microarray experimental data analysis and mining

Author: JiangZuo
Tutor: ZhangShiQiang；BoYouQuan
School: Chongqing Medical University
Course: Biomedical Information Technology
Keywords: Breast Cancer Microarray Significance analysis Correlation Analysis Collaborative filtering
CLC: R737.9
Type: Master's thesis
Year: 2011
Downloads: 98
Quote: 0
Read: Download Dissertation

Abstract

Objective: Breast cancer is a serious hazard to women's health, a higher incidence of cancer, breast cancer awareness from the genetic level pathogenesis mechanisms in cancer research and development has an important significance. Gene chip technology can be automated, large-scale, quickly and easily measure the cancer tissue and adjacent normal tissue gene expression levels, the experimental data obtained by comparing the use of mathematical and computer methods for data analysis and mining is expected to identify presented in different samples and their associated genes differentially expressed genes. At present, a large number of microarray experiments have been published on the Internet, using the Internet can open sharing of these experimental data. This study aimed to be downloaded on the Internet microarray experimental data analysis and mining, mining methods to verify the feasibility and look with breast cancer disease-related genes differentially expressed genes and their associated switching genes for further study Candidate genes and gene regulatory networks for building foundation. Methods: In this study, significance analysis (Significant Analysis of Microarray, SAM) method, the top score genes (Top-Scored Pair, TSP) method to find the cancer tissue and adjacent normal tissues showed differentially expressed genes; using the data Mining association rules (Association Rule) method, collaborative filtering (Collaborative Filtering) method to find a similar or opposite variation of total tune-related genes, switching genes. First, from the search on the Internet to get the raw data of microarray experiments, and then make the necessary data preprocessing, re-use SAM, TSP, association rules, collaborative filtering and other methods for data analysis and mining to identify differentially expressed genes and related genes . Results: In this study, the above method is applied to breast cancer microarray experimental data analysis and mining, to find a number in cancer tissues and adjacent normal tissues showed differential expression of genes, some of which genes have been reported in several papers, this proves that the occurrence and development of breast cancer are closely related; while looking into a number of similar or opposite changes of gene and part of a switching effect of the gene, which, upon inquiry shows some genes in the biological sense indeed correlated genes. Conclusion: The use of SAM and TSP approach to initial screening significantly differentially expressed genes is effective, it can maintain a low false discovery rate, we found significantly greater number of differentially expressed genes; using correlation analysis and collaborative filtering methods to initially find related genes is feasible to identify the genes that are biologically indeed of regulation, and thus have a common genetic variation. Looking into the differentially expressed genes and related genes can be used for further research, and for the initial construction of gene regulatory networks play a fundamental role.

Breast cancer microarray experimental data analysis and mining

Abstract

Related Dissertations