Statistics Colloquium : Dr. Seungchul Baek

Analysis of breast cancer survival data with gene expressions motivates us to use a two-stage approach to model the association between survival time and breast cancer pathway activities. In the first stage, based on five breast cancer subtypes and eight prominent pathways, we identify subtype-specific and pathway-specific informative genes using factor model and estimate factor and loading matrix for each cancer subtype. We further impose a sparsity condition such that prior biological knowledge on relation between pathway activity levels and genes can be incorporated in the model. In the second stage, we investigate the relationship between pathway activity levels and survival time with censoring using a general dimension reduction model in the survival analysis context. Combining the factor model and sufficient direction model provides an efficient way to analyzing high-dimensional data and reveals some interesting relations in the breast cancer gene expression data.