PgmNr P2037: Leveraging haplotype-aware inference for evolve-and-resequence studies.

Authors:
S. Greenblum ¹ ; S. Tilk ¹ ; A. Bergland ^1,2 ; P. Schmidt ³ ; D. Petrov ¹

---

Institutes
1) Stanford University, Stanford, CA; 2) University of Virginia, Charlottesville, VA; 3) University of Pennsylvania, Philadelphia, PA.

Abstract:

In the face of sharp environmental fluctuations, massive adaptive change has been observed in metazoan populations on timescales of just tens of generations. Despite the importance of rapid adaptation for predicting the dynamics of critical health threats such as host-pathogen co-evolution and viral outbreak, the genetic basis of this phenomenon remains unclear.  Large-scale evolve-and-resequence (E+R) experiments, entailing serial sequencing of pooled samples from populations subject to selection, hold enormous promise for studying the dynamics of rapid adaptation in a controlled yet realistic setting.  However, more work is needed to devise experimental and analytic frameworks that maximize the power and precision of these studies, given available technologies, resources, and informatics tools.

Here we leverage both data from the experimental evolution of Drosophila melanogaster populations, as well as simulated data, to demonstrate how various experimental designs incorporating fully-sequenced founder lineages influence the detection resolution of alleles under strong short-term selection.  In particular, we focus on the tradeoff between experimental schemes that minimize linkage disequilibrium among founder haplotypes to uncouple the trajectories of strongly adaptive alleles and nearby sites, and schemes that incorporate new computational tools that make use of founder linkage to increase the precision of individual allele frequency measurements, thereby reducing sequencing costs and facilitating higher levels of replication.

We demonstrate how this tradeoff informs best practices for E+R, and quantify the increased power afforded by linkage-informed analyses in detecting the genomic targets of rapid adaptation.  Improved frameworks building on these results will immediately benefit future studies of experimental evolution in many model organisms, and will help clarify the true extent of the Drosophila genome that is subject to short-term seasonal selection as well as the broader impact of rapid adaptation on genome architecture.