作者: ZHI YAN , JIANGENG LI , YIMIN XIONG , WEITIAN XU , GUORONG ZHENG
DOI: 10.3892/OR.2012.1891
关键词:
摘要: Colon cancer is the third most common and one of leading causes cancer-related death in world. Therefore, identification biomarkers with potential recognizing biological characteristics a key problem for early diagnosis colon patients. In this study, we used random forest approach to discover based on set oligonucleotide microarray data cancer. Real-time PCR was validate related expression levels selected by our approach. Furthermore, ROC curves were analyze sensitivity specificity each biomarker both training test sample sets. Finally, analyzed clinical significance their differential expression. A single classifier consisting 4 genes (IL8, WDR77, MYL9 VIP) forests an average 83.75 76.15%. The validated real-time 48 samples compared matched normal tissues. Patients high IL8 low VIP had significantly reduced median survival rate results indicate that can be employed data. These identified have act as