Project (Anja Sturm)
Stochastic modelling of gene genealogies
We are concerned with describing and analysing gene genealogies in stochastic population models that take into account various complicating factors such as spatial or type structure, selection or diploidy. The goal is to analyse and approximate genealogical relationships of individuals sampled from a present day population. The individuals are represented by single genes (one locus) or gene sequences (multiple loci) that are affected by recombination such that their ancestral relationship is described by an ancestral graph rather than a genealogical tree. From these genealogical relationships coupled with a mutation process we aim to deduce the distribution of quantities that measure the genetic variability in the sample, thereby providing a basis for the analysis of genetic data. We are also interested in using the underlying stochastic models for reconstructing the genealogical relationships from sequence data.
Work on this project requires a background in Mathematics with very good knowledge in probability theory (stochastic process theory). A strong interest in biology and genetics is also a prerequisite.
Publications:
“Coalescent results for diploid exchangeable population models”. With Matthias Birkner and Huili Liu.
Preprint 2017. Available on arxiv.org arxiv:1709.02563.
“On spatial coalescents with multiple mergers in two dimensions". With Benjamin Heuer.
Theoretical population biology, Volume 87, 90-104, 2013.
"The spatial Lambda-coalescent''. With V. Limic. Electronic Journal of Probability,
Volume 11, Number 15, 363-393, 2006.
"Coalescence in a random background''. With N.H. Barton and A.M. Etheridge.
Annals of Applied Probability, Volume 14, Number 2, 754-785, 2004.