Abstract:
|
In the development of genomic biomarkers and molecular diagnostics, high-dimensional genomic studies such as DNA microarray experiments generally require enormous costs and efforts. In clinical and epidemiological studies, several efficient study designs have been developed for reducing the cost of expensive measurements, such as case-cohort designs. Under these efficient designs, expensive measurements are collected only on selected subjects based on their response selective sampling schemes and the measurement costs are effectively reduced. In this study, we discuss to apply these efficient designs to the high-dimensional genomic studies, and formulate statistical methodologies such as gene selection (e.g., multiple testing based on the false discovery rate) and classification/prediction algorithms. In addition, we adapt the efficient semiparametric inference methods to the developments of genomic biomarkers using auxiliary clinical information. Numerical evaluations based on a cancer clinical study are provided.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.