Statistical guidelines for detecting past population shifts using ancient DNA

Tobias Mourier*, Simon Y W Ho, M. Thomas P Gilbert, Eske Willerslev, Ludovic Orlando

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

31 Scopus citations

Abstract

Populations carry a genetic signal of their demographic past, providing an opportunity for investigating the processes that shaped their evolution. Our ability to infer population histories can be enhanced by including ancient DNA data. Using serial-coalescent simulations and a range of both quantitative and temporal sampling schemes, we test the power of ancient mitochondrial sequences and nuclear single-nucleotide polymorphisms (SNPs) to detect past population bottlenecks. Within our simulated framework, mitochondrial sequences have only limited power to detect subtle bottlenecks and/or fast post-bottleneck recoveries. In contrast, nuclear SNPs can detect bottlenecks followed by rapid recovery, although bottlenecks involving reduction of less than half the population are generally detected with low power unless extensive genetic information from ancient individuals is available. Our results provide useful guidelines for scaling sampling schemes and for optimizing our ability to infer past population dynamics. In addition, our results suggest that many ancient DNA studies may face power issues in detecting moderate demographic collapses and/or highly dynamic demographic shifts when based solely on mitochondrial information.

Original languageEnglish (US)
Pages (from-to)2241-2251
Number of pages11
JournalMolecular Biology and Evolution
Volume29
Issue number9
DOIs
StatePublished - Sep 1 2012

Keywords

  • Population bottlenecks
  • ancient DNA
  • serial coalescent

ASJC Scopus subject areas

  • Genetics
  • Molecular Biology
  • Ecology, Evolution, Behavior and Systematics

Fingerprint Dive into the research topics of 'Statistical guidelines for detecting past population shifts using ancient DNA'. Together they form a unique fingerprint.

Cite this