Robust Inference Based on Linear Mixed Models with Application to Gene Detections

• Activity Number: 133 - Gene-Set Based Analysis in Genomic Studies
• Type: Contributed
• Date/Time: Monday, July 30, 2018 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Youngseok Song* and Wen Zhou and Kim Hoke and Wenxin Zhou
• Companies: Colorado State University and Colorado State University and Colorado State University and University of California, San Diego
• Keywords: Robust inference; Linear mixed model; Gene detection; Huber loss

Abstract

Identifying genetically related markers to a candidate pool based on sequencing data has been playing critical roles in genetic and genomics studies for treating cancer, controlling experiment quality, and learning evolutionary pathways. The complex dependency and leptokurtic nature of the sequencing data, however, make the conventional statistical approaches unreliable due to the loss of controlling the false discovery rate or compromising the empirical power. Motivated by [Fan et al., 2017], we consider a factor adjusted linear mixed model by which we can identify genetically related markers through testing linear hypotheses like contrasts. In particular, we propose a robust multiple testing procedure to handle these heavy-tailed high dimensional data. The proposed method can simultaneously resolve the difficulties brought from the inter-gene dependence and heavy-tailedness among data. We demonstrate that our procedure can improve power and control FDP when the data is generated by heavy-tailed distributions both theoretically and numerically. We apply our procedure to Guppy RNA sequencing data from Fischer et al. (2018) and reveal interesting biological insights.