Ruth Shaw

Title: Charlie and evolutionary statistical genetics

Abstract: As Charlie's sister, I was among the first witnesses to Charlie's prowess as a statistician.  Early during my graduate studies, when he had yet to begin his formal statistics studies, I sought his advice on jointly analyzing multiple components of organismal fitness (some binary, others quantitative) to make valid inferences about natural selection. The ideas he sketched then he eventually laid out as forest graph exponential family conditional or unconditional canonical statistic models (which we call 'aster') and implemented them in R.  This advance is widely appreciated by evolutionary biologists. I will survey some of the findings we have gleaned using aster.