Statistics Seminar: "Statistical phylogenetics to infer evolutionary history"

Michael Landis, Washington University in Saint Louis

Abstract: Phylogenetic trees are central to understanding how evolutionary processes generate patterns of biological variation. Increasingly, statistical phylogenetic models are used to infer historical properties of life on earth, such as how species are interrelated, when lineages diverged, and where ancestral species lived. This talk is intended to introduce general problems we biologists encounter in phylogenetics, and their proposed model-based solutions. As a case study, I will discuss recent work of mine concerning the Hawaiian silversword radiation, an adaptive radiation of remarkable island plants that are endemic to a Pacific volcanic island chain. The work demonstrates how evolutionary inferences are improved by jointly modeling the effect of molecular, biogeographical, and paleogeographical events on biodiversity patterns.

Host: Figueroa-Lopez, Jose