Inferring complex demographic histories
There is much interest in analyzing genome-scale DNA sequence data to infer population histories, We developed an efficient, flexible statistical method (diCal 2.0) that can utilize whole-genome sequence data from multiple populations to infer complex demographic models involving population size changes, population splits, admixture, and migration. Together with our collaborators, we applied this method to genomic data from Native American individuals to shed light on the demographic history that underlies the peopling of the Americas. We are currently exploring new ways to develop efficient coalescent-based tools for demographic inference. Relevant publications: Steinrücken, Paul & Song (2013), Steinrücken, Kamm, & Song (2015), Raghavan et. al. (2015), Miroshnikov & Steinrücken (2017).
Inferring strength of selection from time-series
The increased availability of time series genetic variation data from experimental evolution studies and ancient DNA samples has created new opportunities to identify genomic regions under selective pressure and to estimate their associated fitness parameters. We developed a novel spectral method (spectralHMM) to analytically and efficiently integrate over population allele-frequency trajectories underlying the observed temporal DNA data, to compute likelihoods under an evolutionary model. Our method is flexible enough to handle general diploid models of selection. We are currently working on extending these ideas to full genome-sequence data. Relevant publications: Steinrücken, Wang & Song (2013), Steinrücken, Bhaskar & Song (2014), Jewett, Steinrücken & Song (2017).