| 篇名/題名： | AKBs & MKBs: Knowledge-based systems to predict breast cancer mortality |
| 並列篇名： | AKBs 和 MKBs：基於知識的乳癌死亡率預測系統 |
| 摘要： | Breast cancer is a serious health concern that has been significantly impacted by next-generation genomic technologies. In order to comprehend the risk of mortality linked with breast cancer genomic data, a combination of survival analysis and machine learning has been utilized. This study employed logistic regression odds ratios and critical features based on deep learning to extract the level of risk. The odds ratios and risk scores were concatenated into two additional knowledge-based systems using an additive (AKBs) and multiplicative (MKBs) model. Gene set enrichment and gene ontology were utilized to identify gene functions and pathways related to breast cancer mortality. The findings demonstrated that AKBs and MKBs performed better than deep learning and logistic regression models, with an average accuracy of 87%. Next-generation DNA and RNA genomic sequences were used under the clinical characteristics of breast cancer. AKBs and MKBs effectively categorized high-risk missense mutations and candidate genes, which were highly correlated with overall breast cancer mortality. The findings of this study indicated that AKBs and MKBs may improve the identification of candidate genes associated with all-cause mortality. This study introduced two knowledge-based systems (AKBs and MKBs) to identify high-risk missense mutations associated with mortality based on critical features. Our results provide interpretable signals for biomarker discovery in retrospective data; further prospective studies are necessary. |
| 類型： | 期刊論文 |
| 西元出版年： | 2025 |
| 關鍵詞： | Knowledge-based systems
；
Deep learning
；
Mortality
；
Next-generation genomic data |
| 作者： | 作者：
楊正宏
|
| 學校系所： | 資訊管理系 |
| 卷期： | Knowledge-Based Systems；period 333 期 |
