Open Access Open Badges Methodology article

Distinguishing migration from isolation using genes with intragenic recombination: detecting introgression in the Drosophila simulans species complex

Miguel Navascués12, Delphine Legrand34, Cécile Campagne4, Marie-Louise Cariou4 and Frantz Depaulis15*

Author Affiliations

1 UMR 7625 Écologie et Évolution (CNRS/École Normale Supérieure/Université Pierre et Marie Curie), Paris, France

2 INRA, UMR1062 CBGP, F-34988 Montferrier-sur-Lez, France

3 USR 2936, Station d’Écologie Expérimentale du CNRS, 09200 Moulis, France

4 UPR 9034 Évolution, Génomes et Spéciation, CNRS, avenue de la Terrasse, 91198 Gif sur Yvette, France

5 Laboratoire d’Écologie, École Normale Supérieure, 46 rue d’Ulm, 75230 Paris cedex 05, France

For all author emails, please log on.

BMC Evolutionary Biology 2014, 14:89  doi:10.1186/1471-2148-14-89

Published: 24 April 2014



Determining the presence or absence of gene flow between populations is the target of some statistical methods in population genetics. Until recently, these methods either avoided the use of recombining genes, or treated recombination as a nuisance parameter. However, genes with recombination contribute additional information for the detection of gene flow (i.e. through linkage disequilibrium).


We present three summary statistics based on the spatial arrangement of fixed differences, and shared and exclusive polymorphisms that are sensitive to the presence and direction of gene flow. Power and false positive rate for tests based on these statistics are studied by simulation.


The application of these tests to populations from the Drosophila simulans species complex yielded results consistent with migration between D. simulans and its two endemic sister species D. mauritiana and D. sechellia, and between populations D. mauritiana on the islands of the Mauritius and Rodrigues.


We demonstrate the sensitivity of the developed statistics to the presence and direction of gene flow, and characterize their power as a function of differentiation level and recombination rate. The properties of these statistics make them especially suitable for analyzing high-throughput sequencing data or for their integration within the approximate Bayesian computation framework.

Shared polymorphism; Recombination; Hybridization; Gene flow; Incomplete lineage sorting; Drosophila simulans complex