Mixed random forest-based GWAS and a forward-in-time simulation study

I run a statistical genetics lab at UIUC. My lab seeks to explore how statistical genetics can be optimized to gain a better understanding on the genetic architecture of traits. This could lead to even more efficient applications of genomic prediction to crop breeding programs. In this presentation, I am going to give a progress report on both recently completed on three projects. The first involves using a mixed random forest approach to identify rare variants in a manner that rigorously controls the type I error rate. The second is an update on what I presented last year, namely a study of the omnigenic model of genetic architecture. The third is a completed project in which we use empirical distributions of a measurement of prediction accuracy to identify a set of maize genes that putatively also contribute to phenotypic variance in sorghum. I look forward to your feedback, and would be thrilled to incorporate suggestions into our ongoing work.