Data from: On the causes of correlated genomic ancestry across contrasting hybridization histories in a monkeyflower species pair
Data files
Sep 03, 2026 version files 6.30 GB
-
ancestry_calls_raw.txt
203.05 MB
-
ancestry_calls_windowed.txt
8.50 MB
-
ancestry_frequencies_with_features.txt
572.86 KB
-
Ancestry_informative_sites_counts_withcMs.txt
8.91 MB
-
Ancestry_informative_sites.txt
3.90 MB
-
ancestry_sample_summaries_windowed.txt
108.96 KB
-
ancestry_transitions.txt
2.24 MB
-
angsd_genotype_likelihoods.beagle.gz
54.19 MB
-
highqual_sites_FINAL_thinned.pos.txt
1.23 MB
-
highqual_sites_FINAL.pos.txt
55.38 MB
-
panel_counts_highqual_sites.txt
161.55 MB
-
panel_genotypes_filtered_highqual_sites.vcf.gz
5.80 GB
-
PCA_cov_matrix.txt
7.06 MB
-
raw_read_counts.txt
35.14 KB
-
README.md
10.13 KB
-
Southern_population_codes_locations.csv
1.71 KB
Abstract
Hybridization is a powerful force shaping the evolutionary trajectories of species, yet its outcomes are highly variable both across taxa and within a pair of species. In this study, we examine the processes shaping variation in the extent of hybrid ancestry, both among populations and across the genome. We use low-coverage sequencing data to infer local ancestry across the genome for 782 individuals from multiple populations across two geographic regions within the broadly overlapping range of Mimulus guttatus and Mimulus nasutus. We find that the extent of hybrid ancestry is variable across populations, supporting disparate historical patterns of hybridization. However, genomic patterns of hybrid ancestry are correlated across groups, indicating they are shaped by parallel processes. Correlations are highest in geographically proximal populations, including between sympatric and allopatric locations, providing evidence that introgression is not locally constrained but spreads via migration across the landscape. We find that features of the genome are predictive of hybrid ancestry and its correlations among populations. However, contrary to findings in some other species, these patterns are likely not driven by simple linked selection against hybrid ancestry. Genomic outliers for high hybrid ancestry are often shared among populations, suggesting a role for parallel positive selection on ancestry. However, known loci associated with reproductive isolation are poor predictors of ancestry variation across populations, indicating that selection acting in natural hybrid populations is highly polygenic and that the underlying genetic architecture varies across space. Overall, this study demonstrates how ecological, demographic, and genomic features all interact to shape the outcomes of hybridization.
Dataset DOI: 10.5061/dryad.8931zcs5n
Description of the data and file structure
This dataset provides sample-level genotype metadata and processed data related to the analysis of hybrid ancestry across the genome for samples in two regions of contact between the Monkeyflower species Mimulus guttatus and Mimulus nasutus.
Files and variables
File: Ancestry_informative_sites_counts_withcMs.txt
Description: List of ancestry-informative SNP markers used to assess local M. guttatus vs. M. nasutus ancestry, along with allele counts from the reference panel and recombination distances between each marker. File format is that used by the ancestryinfer pipeline. File is tab-separated text with no header. Columns in order are chromosome, position, reference allele count in guttatus panel, alt allele count in guttatus panel, reference allele count in nasutus panel, alt allele count in nasutus panel, recombination rate between each marker and the previous marker.
File: ancestry_sample_summaries_windowed.txt
Description: List of samples, including reference panel, and summary values for genome-wide ancestry based on 50kb-windowed local ancestry inference using AncestryHMM. Each line is one sample.
Variables
- sample: Name of the sample
- n_called: Number of 50kb windows with called ancestry genotypes for this sample
- missing_sites: Number of 50kb windows with missing ancestry genotypes for this sample
- het_sites: Number of 50kb windows with heterozygous ancestry genotypes for this sample
- p2_hom_sites: Number of 50kb windows with homozygous-nasutus ancestry genotypes for this sample
- p2_alleles: Number of nasutus alleles counted (2hom + 1het)
- proportion_p2: Hybrid index -- the proportion of alleles with nasutus ancestry across the genome (p2_alleles/(2*n_called))
- heterozygosity: The proportion of called sites with heterozygous ancestry (het_sites/n_called)
File: Ancestry_informative_sites.txt
Description: List of ancestry-informative sites used to infer local ancestry. File is tab-separated text with no header. Columns are in order: chromosome, position, guttatus (reference) allele, nasutus (alternate) allele
File: ancestry_transitions.txt
Description: List of estimated locations of within-chromosome transitions between ancestry genotypes for each sample. Produced by the ancestryinfer pipeline. Format is tab-separated text with no header. Columns are in order: chromosome, left side of confidence window for transition location, right side of confidence window for transition location, number associated with a sample. Sample numbers in the order of row positions of the ancestry_calls_raw.txt file.
File: Southern_population_codes_locations.csv
Description: Population codes and GPS locations for all collection locations in the Southern region. Northern populations were taken from previously published data.
Variables
- Code: Three-letter population code
- Stands_for: Source of three-letter code name
- Original_name(s): Additional descriptive name of site
- Latitude
- Longitude
- Elevation
- Species_sequenced: 'gut' if only M. guttatus were collected, 'sym' if both M. guttatus and M. nasutus (and possible hybrids) were collected.
File: ancestry_calls_windowed.txt
Description: Ancestry genotype calls for all samples, binned into 50kb windows. First two columns are a window ID (chromosome:window_start_position:window_end_position) and the number of ancestry-informative sites in the window (nsites). Remaining columns each correspond to one sample. Rows each correspond to one 50kb window. Genotypes are 0 for M. guttatus homozygous, 1 for heterozygous, 2 for M. nasutus homozygous, and NA for missing data.
File: raw_read_counts.txt
Description: Number of read pairs sequenced for each sample.
Variables
- file: filename of the forward reads for each sample
- num_seqs: Number of forward reads (equal to number of read pairs) in the raw sequence data.
File: PCA_cov_matrix.txt
Description: Genomic PCA covariance matrix produced by PCangsd. Samples are ordered based on angsd_genotype_likelihoods.beagle.gz file. See PCangsd documentation.
File: angsd_genotype_likelihoods.beagle.gz
Description: Genotype likelihoods for each sample at each site in the angd SNP list, which is thinned to 1 per 1kb. Format is a gzipped, beagle-formatted likelihood file, as produced by angsd using the -GL 2 function. See angsd documentation for details.
File: highqual_sites_FINAL_thinned.pos.txt
Description: List of SNPs used for angsd genotype likelihood analysis, produced by thinning the high-quality sites list to 1 per 1kb. Format is tab-separated text with no header, columns are chromosome and position.
File: ancestry_calls_raw.txt
Description: Ancestry genotype calls for each sample at each ancestry-informative site prior to windowing. Each line is a sample. First column is the sample IDs, additional columns are the sites. Genotypes are 0 for M. guttatus homozygous, 1 for heterozygous, 2 for M. nasutus homozygous, and NA for missing data.
File: panel_counts_highqual_sites.txt
Description: Full list of high-quality filtered sites from the M. guttatus and M. nasutus reference panel, with summary allele counts for each subgroup within the reference panel. ngut = Northern-clade M. guttatus, shut = Southern-clade M. guttatus, nas = M. nasutus, cacgut = M. guttatus from the CAC hybrid site.
Variables
- chrom: chromosome
- pos: position
- ref: reference (M. guttatus) allele
- alt: alternative (M. nasutus) allele
- altcount_ngut: number of alt alleles present in the Northern-clade M. guttatus subgroup
- missing_ngut: number of samples with missing data in the Northern-clade M. guttatus subgroup
- nsamples_ngut: total number of samples in the Northern-clade M. guttatus subgroup
- altcount_nas: number of alt alleles present in the M. nasutus subgroup
- missing_nas: number of samples with missing data in the M. nasutus subgroup
- nsamples_nas: total number of samples in the M. nasutus subgroup
- altcount_sgut: number of alt alleles present in the Southern-clade M. guttatus subgroup
- missing_sgut: number of samples with missing data in the Southern-clade M. guttatus subgroup
- nsamples_sgut: total number of samples in the Southern-clade M. guttatus subgroup
- altcount_cacgut: number of alt alleles present in the M. guttatus CAC-hybrid subgroup
- missing_cacgut: number of samples with missing data in the M. guttatus CAC-hybrid subgroup
- nsamples_cacgut: total number of samples in the M. guttatus CAC-hybrid subgroup
File: highqual_sites_FINAL.pos.txt
Description: Final list of high-quality filtered sites from the reference panel. Format is tab-separated text with no header. Columns are chromosome and position.
File: panel_genotypes_filtered_highqual_sites.vcf.gz
Description: High-quality SNP dataset in variant call format for the M. guttatus and M. nasutus reference panel. Can be opened with open-source tools such as bcftools or GATK.
File: ancestry_frequencies_with_features.txt
Description: Matrix of ancestry frequencies for each sample group, along with genomic features, for each 50kb window. Format is tab-separated text. Ancestry frequencies are the proportion of M. nasutus ancestry alleles within the sample group.
Variables
- chrom: chromosome
- windowend: end position of 50kb genomic window
- M_per_bp_50kb: recombination rate at 50kb resolution, in units of Morgans/bp
- M_per_bp_1Mb: recombination rate binned into 1Mb resolution, in units of Morgans/bp
- ngenes: number of annotated genes within or overlapping the genomic window
- relative_chr_pos: position of the window relative to the ends of the chromosome; scaled so that 0 = chromosome midpoint, 1 = chromosome start or chromosome end.
- CAC_guttatus_137: ancestry frequency for the N-CAC-sym guttatus sample group
- CAC_hybrid_183: ancestry frequency for the N-CAC-sym hybrid sample group
- CAC_nasutus_58: ancestry frequency for the N-CAC-sym nasutus sample group
- LM_hybrid_56: ancestry frequency for the N-LM-sym hybrid sample group
- Southern_allopatric_guttatus_61: ancestry frequency for the S-Foothills-allo guttatus sample group
- Southern_highelev_guttatus_12: ancestry frequency for the S-Montane-sym guttatus sample group
- Southern_highelev_hybrid_16: ancestry frequency for the S-Montane-sym hybrid sample group
- Southern_sympatric_guttatus_174: ancestry frequency for the S-Foothills-sym guttatus sample group
- Southern_sympatric_nasutus_75: ancestry frequency for the S-Foothills-sym nasutus sample group
Code/software
Code associated with this data is available at https://github.com/mfarnitano/Mimulus_hybrid_ancestry. Programs used to produce, process, and analyze this data include:
- bwa (https://github.com/lh3/BWA) for the alignment of raw sequencing reads to the reference genome.
- GATK (https://gatk.broadinstitute.org/hc/en-us) for the genotyping of the reference panel.
- AncestryHMM (https://github.com/russcd/Ancestry_HMM) for the inference of local ancestry.
- angsd (https://www.popgen.dk/angsd/index.php/ANGSD) for genotype likelihood estimation and population structure analyses.
- Most post-processing analysis was conducting using R (https://www.r-project.org) and the tidyverse packages (https://tidyverse.org)
We collected seeds or leaf tissue from individuals from 21 locations across the foothills and mountains of the Sierra Nevada in northern California, USA (Southern region, Figure 1A, C in associated manuscript). Southern region samples were collected in the field from April to June of 2021. We visited most locations multiple times to capture a range of phenological variation, targeting our sampling to obtain both M. nasutus and M. guttatus as well as any likely hybrids or admixed individuals, including sampling both senesced and actively flowering individuals where available. However, we note that sampling was not exhaustive and may not be numerically representative of the proportions of M. guttatus vs. M. nasutus at any one location.
Tissue from wild-growing individuals was dried with silica gel beads in the field. For wild-collected seeds, a subset of seeds from each maternal family were germinated in the UGA Botany greenhouses, where bud or leaf tissue was collected from a single germinant per maternal family and stored at -80 °C until the time of DNA extraction. For all samples, DNA was flash-frozen in liquid nitrogen, ground with metal beads, and extracted using a CTAB protocol with phenol-chloroform extraction (Fishman, 2020). Samples were prepped for low-coverage whole-genome short-read sequencing using a custom tagmentation library prep (Farnitano & Sweigart, 2024) with bulk-prepped Tn5 tagmentation enzyme (Lu et al., 2017), following the same procedure used for earlier CAC and LM samples (Farnitano et al., 2025). Samples were sequenced on an Illumina NovaSeq 6000 machine at the Duke University Sequencing and Genomic Technologies Core Facility with 150bp paired reads to an average depth of ~1.25 million paired reads per sample (median 0.99 million read pairs). We sequenced a total of 382 individuals from the Southern region (Table S1 in associated manuscript).
We combined these Southern region samples with 503 previously sequenced individuals from two Northern sites in Washington, USA: Catherine Creek (CAC), and Little Maui (LM), consisting of M. nasutus, M. guttatus, and admixed individuals (Table S1). CAC individuals were collected as wild tissue samples from across the growing season over three years: 2019, 2021, and 2022; LM individuals were collected only in 2021.