Research ArticleGENETICS

The genomic mosaicism of hybrid speciation

See allHide authors and affiliations

Science Advances  14 Jun 2017:
Vol. 3, no. 6, e1602996
DOI: 10.1126/sciadv.1602996
  • Fig. 1 Population structuring and species information.

    (A) Top: Illustrations of male plumage patterns in house, Italian, and Spanish sparrows modified from Svensson et al. (83). Bottom: A distribution map of house, Italian, and Spanish sparrows throughout Europe and northern Africa (9). (B) PCA of the LD-pruned high-quality SNP set. (C) Population structuring based on admixture analysis for house, Italian, and Spanish populations.

  • Fig. 2 Phylogenetic inference and admixture analysis of the Italian sparrow.

    (A) ADMIXTURE analysis of 100-kb nonoverlapping windows across the Italian sparrows’ chromosomes 1A and Z, with the house sparrow ancestry shown in blue and Spanish sparrow ancestry shown in red. (B) RAxML tree assignment results for 100-kb nonoverlapping windows across the genome for chromosomes 1A and Z. Windows depict whether the Italian sparrows grouped monophyletically with house sparrows (blue), Spanish sparrows (red), in its own clade (green), or were unresolved (white).

  • Fig. 3 Mitochondrial ancestry in the Italian sparrow.

    (A) fastSTRUCTURE analysis of complete mitochondrial sequences for the house, Italian, and Spanish sparrow individuals. (B) Mitochondrial haplotype network of all sparrow individuals. (C) Plots of sliding window (1-kb window and 100-bp step) sequence divergence (dXY) along the mitochondria between the house and Italian (blue) and Spanish and Italian (red) sparrows for three Italian individuals. The top and middle panels depict the mixed mitochondrial ancestry of two Italian individuals, whereas the bottom panel shows the sole house sparrow ancestry of a single Italian sparrow mitochondria that is representable for the eight remaining Italian sparrows (see fig. S3).

  • Fig. 4 Divergence landscape and selection tests for the Passer taxa.

    (A) FST estimates for 100-kb overlapping windows with 25-kb steps across the largest chromosomes in the genome. Windows are identified as HI divergence peaks in blue, SI in red, and PI in yellow. Microchromosomes are plotted in fig. S4. (B) Density plots of TD estimates for PI, SI, and HI outlier windows, as well as all nonoutlier windows (bottom right) for each focal species. (C) Density of DoS values within PI, SI, and HI windows and the genomic background for the three species comparisons. (D) Plots of disparities in FST values between HI/parent species (top) and SI/parent species (second from top) for all genomic windows on the Z chromosome, highlighting windows in PI peaks (yellow), SI peaks (red), and HI peaks (blue). The two bottom panels depict smoothed plots of TD and π in all 100-kb overlapping genomic windows along the Z chromosome for each focal sparrow species (house, blue; Spanish, red; Italian, yellow).

  • Fig. 5 Signatures of selection on the Z chromosome versus autosomes.

    (A) Density of DoS values for all autosomal and Z-linked genes for the HS, SI, and SI species comparisons. (B) Density of TD values for all autosomal and Z-linked genes in the Spanish (red), house (blue), and Italian (yellow) sparrows.

  • Table 1 Mean values of population genomic statistics for 100-kb sliding windows with 25-kb steps across the genome.

    df, density of fixed differences.

    ParameterSpeciesAutosomesZ chromosomeWhole genome
    FSTHS0.32 ± 0.100.45 ± 0.090.33 ± 0.10
    HI0.17 ± 0.070.29 ± 0.160.18 ± 0.08
    SI0.23 ± 0.120.41 ± 0.170.25 ± 0.14
    df (10−3)HS0.048 ± 0.281.440 ± 1.900.146 ± 0.67
    HI0.003 ± 0.010.005 ± 0.010.003 ± 0.01
    SI0.003 ± 0.020.211 ± 0.700.018 ± 0.19
    πHouse0.0073 ± 0.00210.0042 ± 0.00250.0071 ± 0.0023
    Italian0.0070 ± 0.00210.0048 ± 0.00250.0068 ± 0.0023
    Spanish0.0051 ± 0.00220.0025 ± 0.00230.0049 ± 0.0023
    Θ*House0.0075 ± 0.0020.0043 ± 0.0020.0073 ± 0.002
    Italian0.0074 ± 0.0020.0046 ± 0.0020.0072 ± 0.002
    Spanish0.0055 ± 0.0020.0027 ± 0.0020.0053 ± 0.002

    *Watterson estimator of Θ based on the number of segregating sites.

    • Table 2 Results from Patterson’s D test for introgression between lineages with block jack-knifed SE estimates and significance values.
      P1P2P3Patterson’s DJack-knifed SEZ scoreP
      AutosomesHouseItalianSpanish0.1950.00127.77<0.00001
      SpanishItalianHouse0.035<0.0014.98<0.00001
      ZHouseItalianSpanish0.5520.0048.34<0.00001
      SpanishItalianHouse0.5750.0067.62<0.00001
      Whole genomeHouseItalianSpanish0.211<0.00124.23<0.00001
      SpanishItalianHouse0.065<0.0015.39<0.00001
    • Table 3 Mean fd values for 100-kb sliding windows with 25-kb steps across the genome.
      P1P2P3fd (%)
      Autosomes
      fd (%)
      Z chromosome
      fd (%)
      Whole genome
      HouseSpanishItalian31.7 ± 15.440.3 ± 23.732.1 ± 16.0
      SpanishHouseItalian27.5 ± 21.748.7 ± 27.832.4 ± 24.9
    • Table 4 Average population genomic statistics for the high-divergence windows and the genomic background.
      ParameterSpeciesPIHISIBackground
      FSTParents0.3800.4680.5040.329
      HI0.6510.3890.1090.178
      SI0.7030.1250.5740.241
      TDHouse−1.5160.0100.292−0.132
      Italian1.625−1.0070.750−0.206
      Spanish−0.979−0.785−0.576−0.374
      πHouse0.00140.00460.00570.0072
      Italian0.00380.00300.00570.0069
      Spanish0.00080.00230.00210.0050
      r2*House0.44150.49010.44270.3155
      Italian0.55270.43090.39570.2658
      Spanish0.30710.38960.33950.2736
      ΘHouse0.00210.00460.00540.0073
      Italian0.00290.00370.00500.0073
      Spanish0.00100.00270.00240.0054

      *Mean estimates of LD in the form of pairwise r2 estimates for all SNPs within 1 kb of each other in the specified genomic windows.

      †Watterson’s estimator of Θ based on the number of segregating sites.

      • Table 5 Significantly enriched functional groups within outlier windows.
        Window typeFunctional groupBonferroni step-down corrected P
        HIAmino acid transport9.0995 × 10−4
        HINegative regulation of G protein–coupled receptor protein signaling pathway1.6878 × 10−4
        HIPositive regulation of translation0.0492
        HIRegulation of guanosine triphosphatase activity1.5561 × 10−5
        HIRegulation of autophagy0.0012
        HIRegulation of muscle system process1.4758 × 10−5
        SIBehavioral response to nicotine1.1522 × 10−4
        SICell differentiation in the spinal cord0.0026
        SICell differentiation involved in kidney development0.0016
        SIErythrocyte homeostasis0.0205
        SIPalate development0.0201
        SIRegulation of BMP signaling pathway0.0257
        SIRegulation of mRNA processing0.0069
        SIRegulation of organ growth0.0116
        PICellular response to retinoic acid0.0042
        PIDorsal/ventral pattern formation0.0027
        PINegative regulation of innate immune response0.0081
        PINegative regulation of stress-activated mitogen-activated protein kinase cascade0.0027
        PIRegulation of anion transmembrane transport0.0017

      Supplementary Materials

      • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/6/e1602996/DC1

        fig. S1. Stepping window ADMIXTURE and RAxML analyses across the Italian sparrow genome.

        fig. S2. LD decay across all chromosomes.

        fig. S3. Mitochondrial sequence divergence between each Italian sparrow and the parent populations.

        fig. S4. Divergence peaks between the Italian sparrow and either parent taxa on microchromosomes.

        fig. S5. Divergence peaks for PI regions on large chromosomes.

        fig. S6. Population structuring analysis of sparrow individuals.

        table S1. Sequencing scheme for the house sparrow reference genome assembly.

        table S2. Final assembly statistics.

        table S3. Sample information.

        table S4. Results from genome-wide RAxML analysis.

        table S5. Comparison of resolved RAxML phylogenies at variable window sizes.

        table S6. ENSEMBLE gene IDs for genes within top divergence windows.

        table S7. Significantly enriched GO pathways from ClueGo analysis in HI outlier windows.

        table S8. Significantly enriched GO pathways from ClueGo analysis in SI outlier windows.

        table S9. Significantly enriched GO pathways from ClueGo analysis in PI outlier windows.

        table S10. Genes among outlier windows identified as candidates for the involvement in melanogenesis and RI in sparrows.

        table S11. Significantly enriched GO pathways from ClueGo analysis HI outlier windows with chickens as the reference genome.

        table S12. Significantly enriched GO pathways from ClueGo analysis in SI outlier windows with chickens as the reference genome.

        table S13. Significantly enriched GO pathways from ClueGo analysis in PI outlier windows with chickens as the reference genome.

      • Supplementary Materials

        This PDF file includes:

        • fig. S1. Stepping window ADMIXTURE and RAxML analyses across the Italian sparrow genome.
        • fig. S2. LD decay across all chromosomes.
        • fig. S3. Mitochondrial sequence divergence between each Italian sparrow and the parent populations.
        • fig. S4. Divergence peaks between the Italian sparrow and either parent taxa on microchromosomes.
        • fig. S5. Divergence peaks for PI regions on large chromosomes.
        • fig. S6. Population structuring analysis of sparrow individuals.
        • table S1. Sequencing scheme for the house sparrow reference genome assembly.
        • table S2. Final assembly statistics.
        • table S3. Sample information.
        • table S4. Results from genome-wide RAxML analysis.
        • table S5. Comparison of resolved RAxML phylogenies at variable window sizes.
        • table S6. ENSEMBLE gene IDs for genes within top divergence windows.
        • table S7. Significantly enriched GO pathways from ClueGo analysis in HI outlier windows.
        • table S8. Significantly enriched GO pathways from ClueGo analysis in SI outlier windows.
        • table S9. Significantly enriched GO pathways from ClueGo analysis in PI outlier windows.
        • table S10. Genes among outlier windows identified as candidates for the involvement in melanogenesis and RI in sparrows.
        • table S11. Significantly enriched GO pathways from ClueGo analysis HI outlier windows with chickens as the reference genome.
        • table S12. Significantly enriched GO pathways from ClueGo analysis in SI outlier windows with chickens as the reference genome.
        • table S13. Significantly enriched GO pathways from ClueGo analysis in PI outlier windows with chickens as the reference genome.

        Download PDF

        Files in this Data Supplement: