|
Identifying FCN2 as a Potential Biomarker for Hepatocellular Carcinoma Using a Novel Supervised Classification Method |
|---|---|
| รหัสดีโอไอ | |
| Creator | Nuttachat Wisittipanit, Ekachai Chukeatirote |
| Title | Identifying FCN2 as a Potential Biomarker for Hepatocellular Carcinoma Using a Novel Supervised Classification Method |
| Contributor | Nuttachat Wisittipanit, Ekachai Chukeatirote |
| Publisher | Genetics Society of Thailand |
| Publication Year | 2562 |
| Journal Title | Genomics and Genetics |
| Journal Vol. | 12 |
| Journal No. | 1 |
| Page no. | 19-27 |
| Keyword | hepatocellular carcinoma, gene expression, machine learning, attribute selection, genetic marker |
| URL Website | https://www.tci-thaijo.org/index.php/gst/issue/view/13960 |
| Website title | https://www.tci-thaijo.org/index.php/gst/article/view/147436 |
| ISSN | 24655198 |
| Abstract | Hepatocellular Carcinoma (HCC) is the most prevalent form of liver cancers of which most patients show no symptoms of the condition until the ailment is beyond the curable stage. Computational analysis using data mining methods were employed to analyze gene expression profiles on microarray chips from 511 samples (268 HCC tissues and 243 adjacent non-HCC tissues) to identify highly differentially expressed genes. Of 23,000 genes available in the platform, 500 genes (~2%) were preliminary identified by Relief algorithm based on the high ranking scores of their expression levels between HCC-infected tissues and normal tissues. Then, the intensive search for differentially expressed genes were performed by machine learning methods; namely Support Vector Machines (SVM), K-Nearest Neighbour (KNN) and Na?ve Bayes Classification (NBC) algorithm, by attribute pool sweep process. Results showed that several highly differentially expressed genes were detected having relatively high classification accuracy with Ficolin-2 acting as a common discovered gene. |
