Data from; Vegetative propagation preserves genomic diversity and informs translocation strategies in a rare clonal plant
Abstract
This dataset contains genome-wide SNP data and demographic-model input files used to evaluate genetic diversity, relatedness, population structure, and translocation outcomes in Pleuropogon oregonus, a rare clonal grass endemic to eastern Oregon, USA. The dataset includes two VCF files generated with Stacks v2.65 from 138 sampled individuals representing natural and introduced P. oregonus sites, plus Pleuropogon refractus outgroup samples. One VCF contains all filtered SNPs per locus, while the second retains one randomly selected SNP per locus for analyses requiring unlinked markers. Data values include SNP genotypes, allele depths, read depths, genotype likelihoods, genotype quality scores, allele frequencies, and sample identifiers. Associated log files document filtering parameters and run summaries. The dataset also includes fastsimcoal2 input files for two demographic scenarios, strict isolation and isolation with migration, used to infer divergence history between northern and southern P. oregonus lineages from folded joint site-frequency spectra. These files can be reused for population genomic analyses, conservation genetic comparisons, demographic inference, evaluation of clonal diversity, and method development involving RADseq-derived SNP datasets. They may also support future conservation planning by enabling comparison with additional samples from natural or introduced populations. Plant material was collected under appropriate permits and access permissions. The dataset contains no human subjects data, but because P. oregonus is rare and conservation-sensitive, reuse should avoid facilitating disturbance to natural populations or disclosure of more precise locality information than is already provided in the associated publication.
Dataset overview
This dataset contains genome-wide SNP data and demographic-model input files used to evaluate genetic diversity, relatedness, population structure, and translocation outcomes in Pleuropogon oregonus, a rare clonal grass endemic to eastern Oregon, USA. The dataset also includes samples of Pleuropogon refractus, which were used as an outgroup in population genomic analyses.
The Dryad release is organized as two compressed archives:
vcf.zip: contains Variant Call Format (VCF) files and their associated Stacks log filesFSC2.zip: contains input files for two fastsimcoal2 demographic models
The SNP datasets were generated from ddRADseq data using Stacks v2.65. The VCF archive contains one file with all filtered SNPs per RAD locus and one file with a single randomly selected SNP per locus. The all-SNP dataset was used for genetic summary statistics and pairwise differentiation analyses, while the one-SNP-per-locus dataset was used for analyses requiring reduced linkage among markers, including population structure and demographic analyses.
The fastsimcoal2 archive contains files for two demographic scenarios comparing divergence between northern and southern P. oregonus lineages: a strict-isolation model and an isolation-with-migration model. These files provide model templates, parameter priors, and observed site-frequency spectra needed to rerun or modify the demographic analyses.
Plant material was collected under appropriate permits and access permissions. The dataset contains no human subjects data. Because P. oregonus is rare and conservation-sensitive, reuse should avoid facilitating disturbance to natural populations or disclosure of more precise locality information than is provided in the associated publication.
Top-level Dryad files
dryad/
|-- README.md
|-- vcf.zip
`-- FSC2.zip
vcf.zip
- Archive size: 50,916,882 bytes
- Contents: 4 files
- Purpose: SNP genotypes in VCF format and logs documenting the Stacks
populationsruns that generated them
Archive contents:
vcf.zip
|-- populations_1SNP.log
|-- populations_1SNP.vcf
|-- populations_allSNP.log
`-- populations_allSNP.vcf
FSC2.zip
- Archive size: 9,769 bytes
- Contents: fastsimcoal2 model directories and files
- Purpose: input files for strict-isolation and isolation-with-migration demographic scenarios
Data-bearing archive contents:
FSC2.zip
`-- FSC2/
|-- PLOR_IM/
| |-- PLOR_IM.est
| |-- PLOR_IM.tpl
| |-- PLOR_IM_MSFS.obs
| `-- PLOR_IM_jointMAFpop1_0.obs
`-- PLOR_SI/
|-- PLOR_SI.est
|-- PLOR_SI.tpl
|-- PLOR_SI_MSFS.obs
`-- PLOR_SI_jointMAFpop1_0.obs
The FSC2.zip archive may also contain macOS metadata entries such as .DS_Store, __MACOSX/, or files beginning with ._. These are not data files and can be ignored.
Files in vcf.zip
After extracting vcf.zip, the archive contains two uncompressed VCFv4.2 files generated by the populations program in Stacks v2.65 on 2024-09-19. Each VCF contains 138 sample columns. Sample identifiers are the identifiers used during analysis and include the following prefixes: 7mi, camas, horse, LogV, lv, mud, odot, and refrac. See the associated publication and supplemental materials for the relationship between sample identifiers, sampling sites, natural populations, introduction sites, and the outgroup taxon.
populations_allSNP.vcf
This file contains all filtered SNPs retained by Stacks across RAD loci.
- Uncompressed file size: 166,261,771 bytes
- Lines: 60,357
- Samples: 138
- Variant records: 60,342
- RAD loci/contigs represented: 27,794
- Maximum alternate alleles per record: 1
- Stacks command: generated without
--write-random-snp - Primary use in the associated study: site-level genetic summary statistics and pairwise FST analyses
The corresponding Stacks log is populations_allSNP.log.
populations_1SNP.vcf
This file contains one randomly selected SNP per RAD locus, generated using the Stacks --write-random-snp option.
- Uncompressed file size: 78,349,945 bytes
- Lines: 27,809
- Samples: 138
- Variant records: 27,794
- RAD loci/contigs represented: 27,794
- Maximum alternate alleles per record: 1
- Stacks command: generated with
--write-random-snp - Primary use in the associated study: analyses requiring reduced linkage among loci, including PCA, STRUCTURE, and folded joint site-frequency spectrum construction for fastsimcoal2
The corresponding Stacks log is populations_1SNP.log.
VCF field descriptions
The VCF files contain the standard fixed VCF fields:
#CHROM: Stacks catalog locus identifierPOS: position of the SNP within the locusID: variant identifier, formatted as locus and site informationREF: reference alleleALT: alternate alleleQUAL: variant quality fieldFILTER: filter statusINFO: site-level annotationsFORMAT: genotype-level data fields- Sample columns: genotype and associated values for each sampled individual
The VCF headers define the following INFO fields:
AD: total depth for each alleleAF: alternate allele frequencyDP: total read depthNS: number of samples with dataloc_strand: genomic strand on which the corresponding Stacks locus aligns
The genotype records use the following FORMAT field order:
GT:DP:AD:GQ:GL
These genotype-level fields are:
GT: called genotypeDP: read depth for the individual at the SNPAD: allele depths for the reference and alternate allelesGQ: genotype qualityGL: genotype likelihoods
Missing genotypes are encoded in standard VCF format as ./..
Stacks log files in vcf.zip
populations_allSNP.log
This log records the Stacks populations run used to generate populations_allSNP.vcf.
Important settings and outcomes include:
- Stacks version: v2.65
- Samples in population map: 138
- Populations in population map: 8
- Minor allele frequency cutoff: 0.02
- Maximum observed heterozygosity cutoff: 0.7
- FST correction: p-value correction
- Loci retained after sample/population constraints: 364,245
- Sites represented across retained loci: 40,079,906
- Filtered sites: 16,720
- Variant sites retained: 60,342
The command recorded in the log did not include --write-random-snp, so multiple SNPs per RAD locus could be retained.
populations_1SNP.log
This log records the Stacks populations run used to generate populations_1SNP.vcf.
Important settings and outcomes include:
- Stacks version: v2.65
- Samples in population map: 138
- Populations in population map: 8
- Minor allele frequency cutoff: 0.02
- Maximum observed heterozygosity cutoff: 0.7
- FST correction: p-value correction
- Loci retained after sample/population constraints: 364,245
- Sites represented across retained loci: 40,079,906
- Filtered sites: 16,720
- Variant sites retained: 27,794
The command recorded in the log included --write-random-snp, so one SNP was randomly selected per RAD locus.
Files in FSC2.zip
After extracting FSC2.zip, the FSC2/ directory contains input files for demographic inference in fastsimcoal2 v2.8. These files compare two simplified divergence scenarios between northern and southern P. oregonus lineages using folded joint site-frequency spectra derived from the one-SNP-per-locus dataset.
The two model directories are:
FSC2/PLOR_IM/: isolation-with-migration modelFSC2/PLOR_SI/: strict-isolation model
Each model directory contains four data files:
.tpl: fastsimcoal2 template file describing the demographic model.est: parameter file specifying priors, distributions, and derived parameters_MSFS.obs: observed multidimensional site-frequency spectrum in fastsimcoal2 format_jointMAFpop1_0.obs: observed joint minor allele frequency spectrum in table format
The .obs files are the same in both model directories because both models are fit to the same observed data. The .tpl and .est files differ because they define different demographic scenarios.
FSC2/PLOR_IM/
This directory contains files for the isolation-with-migration model. In this model, northern and southern lineages diverge at time TDIV, and post-divergence migration can occur between demes.
Files:
PLOR_IM.tpl: model template with two demes, sample sizes of 6 and 16 gene copies, one historical divergence event, and migration parametersMIG1andMIG2PLOR_IM.est: parameter priors forANCSIZE,NPOP1,MIG1,MIG2, andTDIV; also definesRESIZE = ANCSIZE/NPOP1PLOR_IM_MSFS.obs: observed folded multidimensional site-frequency spectrumPLOR_IM_jointMAFpop1_0.obs: observed joint minor allele frequency spectrum
Parameters in PLOR_IM.est:
ANCSIZE: ancestral population size, sampled from a uniform distribution from 100 to 500,000NPOP1: contemporary population size for population 1, sampled from a uniform distribution from 1 to 100,000MIG1: migration rate parameter, sampled from a log-uniform distribution from 1e-12 to 1e-1MIG2: migration rate parameter, sampled from a log-uniform distribution from 1e-12 to 1e-1TDIV: divergence time, sampled from a uniform distribution from 100 to 1,000,000RESIZE: derived parameter calculated asANCSIZE/NPOP1
FSC2/PLOR_SI/
This directory contains files for the strict-isolation model. In this model, northern and southern lineages diverge at time TDIV without post-divergence migration.
Files:
PLOR_SI.tpl: model template with two demes, sample sizes of 6 and 16 gene copies, one historical divergence event, and migration matrices fixed to zeroPLOR_SI.est: parameter priors forANCSIZE,NPOP1, andTDIV; also definesRESIZE = ANCSIZE/NPOP1PLOR_SI_MSFS.obs: observed folded multidimensional site-frequency spectrumPLOR_SI_jointMAFpop1_0.obs: observed joint minor allele frequency spectrum
Parameters in PLOR_SI.est:
ANCSIZE: ancestral population size, sampled from a uniform distribution from 100 to 500,000NPOP1: contemporary population size for population 1, sampled from a uniform distribution from 1 to 100,000TDIV: divergence time, sampled from a uniform distribution from 100 to 1,000,000RESIZE: derived parameter calculated asANCSIZE/NPOP1
Reuse notes
To use the VCF data, extract vcf.zip and analyze the uncompressed VCF files with tools such as VCFtools, bcftools, PLINK, R packages that read VCF data, or custom scripts. Users should apply filters appropriate to their analysis. In the associated study, additional downstream filtering was applied for some analyses, including removal of SNPs with greater than 10 % missing data for population structure analyses.
To use the fastsimcoal2 inputs, extract FSC2.zip and run fastsimcoal2 from within the relevant model directory or provide the appropriate file prefix to fastsimcoal2. These files can be reused to reproduce the two demographic models, adjust priors or model assumptions, or compare additional demographic scenarios using the same observed spectra. Users rerunning fastsimcoal2 should check that file names and prefixes match the command-line model prefix used in their run.
Software referenced
-
Stacks v2.65
-
fastsimcoal2 v2.8
-
VCFtools
-
STRUCTURE v2.3.4
-
R and associated population genomic packages used in the associated publication