KCC 2021
Current Result Document :
| 한글제목(Korean Title) |
유방암 전이 관련 핵심 유전자를 이용한 기계학습 예측 모델 |
| 영문제목(English Title) |
Machine learning for prediction of breast cancer metastasis using hub gene |
| 저자(Author) |
김병철
김진규
김강산
우상근
Byung-chul Kim
Jingyu Kim
Kangsan Kim
Sang-Keun Woo
|
| 원문수록처(Citation) |
VOL 48 NO. 01 PP. 0822 ~ 0824 (2021. 06) |
한글내용
(Korean Abstract) |
In this study, we have investigated a new approach to predict of metastasis of breast cancer thought machine learning algorithm with next generation sequencing (NGS) analysis. The NGS data was obtained in TCGA/TCIA database. We downloaded 96 patient’s data of breast cancer, each sample was comprised of 24 metastasis condition (M1 stage) and 72 non-metastasis condition (M0 stage For the more elaborate assay, the gene modules were calculated by GSEA (ver4.0.3). ㄸ Each gene was separated to the modules by gene expression pattern and gene function. The machine learning classification methods were Logistic Regression (LR), Linear Discriminant Analysis (LDA), K-Nearest Neighbor (KNN), Classification and Regression Trees (CART), Naive Bayes (NB), Support Vector Machine (SVM), Random Forest (RF) and Gradient Boosting (GB). The samples were randomly divided into a training and testing at 7:3 ratio. we obtained a two major functions. After differential assay, 448 genes were identified as up-regulated for breast cancer metastasis and 1,055 genes were down-regulated. 14 gene modules were separated by pattern of gene expression, two modules were selected wit function related to breast cancer metastasis. 79 genes were used for estimation of machine learning from Fatty acid metabolism and oxidative phosphorylation related gene modules. The RF model and GB model performed better than the other two algorithm model. In hub gene group, the GB model accuracy value was the highest at 0.88. In hub gene group, the RF model accuracy value was the highest at 0.89. This study suggests that predicting metastasis of breast cancer with NGS will give us to useful information for treat |
영문내용
(English Abstract) |
|
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
