How Tseng
Research Subjects
How Tseng has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Bao Tu Thai Nguyen
Trường Đại học Y dược Cần Thơ
1 shared publication - Duy Nguyen Anh Tran
Can Tho University of Medicine and Pharmacy
1 shared publication - Nguyen Quoc Khanh Le
Taipei Medical University
1 shared publication - Nguyen-Kieu Viet-Nhi
International Master/Ph.D. Program in Medicine, College of Medicine, Taipei Medical University, Taipei 110, Taiwan
1 shared publication - Shih-Han Hung
Department of Otolaryngology, Mennonite Christian Hospital, Hualien, Taiwan
1 shared publication - Thi Thuy Nguyen
International Ph.D. Program in Medicine, College of Medicine, Taipei Medical University, Taipei 110, Taiwan
1 shared publication
Latest Publications
- Open Access
- 3 September 2023
- Published by Association for Computing Machinery (ACM)
- p. 1
Predicting Tumor Mutational Burden and Survival in Head and Neck Squamous Cancer Patients Using Machine Learning and Bioinformatics Approaches
Head and neck squamous cancer (HNSC) is a prevalent malignancy with a complex genetic profile. Tumor Mutational Burden (TMB) is an emerging biomarker associated with prognostic and therapeutic implications. In this study, we aimed to develop machine-learning models for predicting TMB and patient survival using RNA sequencing data and clinical features. Methods: We collected RNA sequencing data and clinical information of HNSC patients from Gene Expression Omnibus (GEO) (GSE142083) and The Cancer Genome Atlas (TCGA) databases. Machine learning models, including Random Forest, Support Vector Machine (SVM), k-Nearest Neighbors (kNN), and Logistic Regression, were built to predict TMB levels based on patient gene expression profiles. The top 100 important features (genes) were selected from these models to create a survival prediction model. Results: Among the tested models, Random Forest showed the highest accuracy (0.8011), followed by SVM (0.796), kNN (0.777), and logistic regression (0.704). Using the top 100 important genes, we developed a model to predict HNSC patient survival (under 3 years, 3--5 years, and over 5 years). Random forest achieved an accuracy of 0.70, while SVM and kNN reached 0.65. We identified five genes (KRT14, KRT6B, COL1A1, FN1, KRT6C) most closely related to TMB and patient survival. Through KEGG pathway analysis and neural network approaches, we discovered that these genes play a significant role in three pathways: PI3K-Akt signaling pathway, Human papillomavirus infection, and Bacterial invasion of epithelial cells. In Conclusion, our study highlights the potential of machine learning in integrating bioinformatics for predicting TMB and patient survival in HNSC. The identified genes (KRT14, KRT6B, COL1A1, FN1, KRT6C) and related pathways may serve as potential biomarkers and therapeutic targets in HNSC treatment and prognosis.
- 3 September 2023
- Published by Association for Computing Machinery (ACM)
- p. 1
Open AccessPredicting Tumor Mutational Burden and Survival in Head and Neck Squamous Cancer Patients Using Machine Learning and Bioinformatics Approaches
Head and neck squamous cancer (HNSC) is a prevalent malignancy with a complex genetic profile. Tumor Mutational Burden (TMB) is an emerging biomarker associated with prognostic and therapeutic implications. In this study, we aimed to develop machine-learning models for predicting TMB and patient survival using RNA sequencing data and clinical features. Methods: We collected RNA sequencing data and clinical information of HNSC patients from Gene Expression Omnibus (GEO) (GSE142083) and The Cancer Genome Atlas (TCGA) databases. Machine learning models, including Random Forest, Support Vector Machine (SVM), k-Nearest Neighbors (kNN), and Logistic Regression, were built to predict TMB levels based on patient gene expression profiles. The top 100 important features (genes) were selected from these models to create a survival prediction model. Results: Among the tested models, Random Forest showed the highest accuracy (0.8011), followed by SVM (0.796), kNN (0.777), and logistic regression (0.704). Using the top 100 important genes, we developed a model to predict HNSC patient survival (under 3 years, 3--5 years, and over 5 years). Random forest achieved an accuracy of 0.70, while SVM and kNN reached 0.65. We identified five genes (KRT14, KRT6B, COL1A1, FN1, KRT6C) most closely related to TMB and patient survival. Through KEGG pathway analysis and neural network approaches, we discovered that these genes play a significant role in three pathways: PI3K-Akt signaling pathway, Human papillomavirus infection, and Bacterial invasion of epithelial cells. In Conclusion, our study highlights the potential of machine learning in integrating bioinformatics for predicting TMB and patient survival in HNSC. The identified genes (KRT14, KRT6B, COL1A1, FN1, KRT6C) and related pathways may serve as potential biomarkers and therapeutic targets in HNSC treatment and prognosis.