Files
Download Full Text (419 KB)
Abstract
Background: Head and neck cancer is the 6th most common cancer worldwide with an expected 1.08 million new cases each year. Such cancer data are ultra-high dimensional with thousands of clinical features and gene expressions, making it challenging for the traditional analytical tools to extract the potential biomarker for the cancer survival and control false discoveries. In addition, presence of heavy censoring can affect the screening procedures based on Kaplan-Meier (K-M) survival estimates.
Aim: To propose a model free, ultra-high dimensional feature screening method with two-dimensional survival outcome allowing false discovery rate (FDR) control.
Method: 516 primary tumor patients with 17702 normalized mRNA gene expressions & 16 clinical covariates were analyzed. covariates associated with survival were identified using a screening procedure. Further, an FDR control procedure was introduced along with the screening algorithm. Performance was evaluated using cross validated AUC values and log-rank test.
Results: A total of 12 covariates (8 gene expressions and 4 clinical features) were selected in the screening with FDR control procedure. AUC values of 1-,3-, 5- years overall survival were 0.75,0.71 and 0.68 respectively. The selected covariates also significantly differentiated low and high risk patients (p-value <.001)
Conclusion: The proposed method can capture both linear and non-linear correlation between predictors and outcome without requiring any complex estimation or optimization. Furthermore, the issue of false discoveries in the screening procedure was also tackled. Thus, the method can be used to identify potential risk factors to improve early screening, treatment and ensure prolong survival.
Publication Date
2023
Keywords
survival analysis, variable screening, FDR control, ultra-high dimension, head and neck cancer
Disciplines
Biostatistics | Genomics | Microarrays | Statistical Models | Survival Analysis
Faculty Advisor/Mentor
Dr. Dipankar Bandyopadhyay, Dr. Chenlu Ke.
Is Part Of
VCU Graduate Research Posters
Included in
Biostatistics Commons, Genomics Commons, Microarrays Commons, Statistical Models Commons, Survival Analysis Commons