A current debate within population genomics surrounds the relevance of patterns of genomic differentiation between closely related species for our understanding of adaptation and speciation. Mounting evidence across many taxa suggests that the same genomic regions repeatedly develop elevated differentiation in independent species pairs. These regions often coincide with high gene density and/or low recombination, leading to the hypothesis that the genomic differentiation landscape mostly reflects a history of background selection, and reveals little about adaptation or speciation. A comparative genomics approach with multiple independent species pairs at a timescale where gene flow and ILS are negligible permits investigating whether different evolutionary processes are responsible for generating lineage-specific versus shared patterns of species differentiation. We use whole-genome re-sequencing data of 195 individuals from four Ficedula flycatcher species comprising two independent species pairs: collared and pied flycatchers, and red-breasted and taiga flycatchers. We found that both shared and lineage-specific FST peaks could partially be explained by selective sweeps, with recurrent selection likely to underlie shared signatures of selection, while indirect evidence supports a role of recombination landscape evolution in driving lineage specific-signatures of selection. This work therefore provides evidence for an interplay of positive selection and recombination to genomic landscape evolution.

This dataset contains SNP calls generated with GATK HaplotypeCaller followed by GenotypeGVCFs for collared, pied, red-breasted and taiga flycatchers, as well as the outgroup Ficedula hyperythra. Hard filter thresholds have been applied to the data (QD < 2.0, FS > 60.0, MQ < 40.0, MQRankSum < -12.5, ReadPosRankSum < -8.0), and SNPs falling within annotated repeats are removed. 9 pied and 7 taiga samples with high amounts of missing data have been removed. Genotype filters for minimum depth of 5, maximum depth 200 and minimum genotype quality 30 have been applied, however sites with high amounts of missing datahave not yet been removed from this file. Only scaffolds mapping to autosomes according to the collared flycatcher reference genome are included.

Additional files contain positions of SNPs with more than 10 percent missing data for each of the species or species pairs. For species pairs, sites with greater than 10% missing data in either species were removed. 

A bed file contains the coordinates of regions identified as putatively collapsed duplicates in the assembly. SNPs falling within these regions were removed from all analyses.