Publications by authors named "Hyoseok Ko"
Multiple Group Testing Procedures for Analysis of High-Dimensional Genomic Data.
Genomics Inform
December 2016
Article Synopsis
- The text discusses the challenges of identifying genes related to diseases using high-dimensional genomic data, highlighting issues like multiple testing errors and the need for effective group testing procedures.
- It compares traditional statistical group testing methods (like PCA and Hotelling's T test) with a regularization approach known as group lasso, which uses penalized likelihood for regression analysis on genomic markers.
- The study found significant discrepancies in the genes identified as associated with ovarian cancer between the traditional group testing methods and the group lasso approach, suggesting that different methods yield different results in gene selection.