Skip to main content

On the genetic architecture of rapidly adapting and convergent life history traits in guppies

Cite this dataset

Whiting, James et al. (2022). On the genetic architecture of rapidly adapting and convergent life history traits in guppies [Dataset]. Dryad.


The genetic basis of traits shapes and constrains how adaptation proceeds in nature; rapid adaptation can be facilitated by polygenic traits, which subsequently provide multiple, redundant, genetic routes to adaptive phenotypes, reducing re-use of the same genes (genetic convergence). Guppy life history traits evolve rapidly and convergently among natural high- (HP) and low-predation (LP) environments in northern Trinidad. This system has been studied extensively at the phenotypic level, but little is known about the underlying genetic architecture. Here, we use an F2 QTL design to examine the genetic basis of seven (five female, two male) guppy life history phenotypes to assess whether the genetic architecture of these traits reflects theoretical predictions. We use RAD-sequencing data (16,539 SNPs) from 370 male and 267 female F2 individuals. We perform linkage mapping, estimates of genome-wide and per-chromosome heritability (multi-locus associations), and QTL ma pping (single-locus associations). Our results are consistent with architectures of many-loci of small effect for male age and size at maturity and female interbrood period. Male trait associations are clustered on specific chromosomes, but female interbrood period exhibits a weak genome-wide signal suggesting a potentially highly polygenic component. Offspring weight and female size at maturity are also associated with a single significant QTL each. These results suggest rapid phenotypic evolution of guppies may be facilitated by polygenic trait architectures, but these could fuel redundancy and limit gene re-use across populations, in agreement with an absence of strong signatures of genetic convergence from recent population genomic analyses of wild HP-LP guppies.


Sequencing data was derived through RAD-sequencing of four F2 cross families (F0s and F2s sequenced).

Phenotype data was derived by phenotyping lab-reared individuals according to the methods in Whiting et al. 2022.

The linkage map was made using LepMap3.

Usage notes

This repository contains data associated with the study "On the genetic architecture of rapidly adapting and convergent life history traits in guppies" by Whiting et al. 2022

Included are four data files:

  * `guppy_LH_qtl_RAD_snps.vcf.gz` - 16,539 RADseq genotypes for 661 individual guppies from four F2 crosses. This is the primary SNP data for this study, from which all marker sets are derived.

  * `qtl_female_phenotypes_clean_with_litters.csv` - Phenotypes for females

  * `qtl_male_phenotypes_clean.csv` - Phenotypes for males

  * `linkage_map_guppy_LH_final.txt` - Final linkage map used


* Female phenotypes include age at first brood (days), size at first brood (mm), first brood size (number of offspring), interbrood interval (days), and dry offspring weight (grams) of brood 1 and brood 2 (only offspring weight from the first brood was used due to correlation/redundancy of these two phenotypes).

* Dry offspring weight was calculated as the average weight of an individual, so (total dry brood weight)/(brood size)

* Male phenotpyes include age at maturity (days), and size (mm) and weight (grams) at maturity (only size was used for analyses due to high correlation between size and weight).

* In each .csv, the "ID" column corresponds to the number in the middle of the IDs of the VCF file, for e.g. "09_556.sorted.rg" in the VCF corresponds to ID 556.

* Missing data is the male and female files are denoted as NA. These are individuals that were sequenced but weren't phenotyped, and were excluded from QTL analyses as described in the manuscript.

* The linkage map includes the original marker number "marker" (1-16539, corresponding to order in VCF); the cM position on lepmap-derived linkage groups "cM"; the original VCF mapped position in bp "mapped"; the original chromosome from the vcf "original_chr"; the lepmap-derived linkage group "LG"; and a column called "trim" that was used to trim the full linkage map ends but can be ignored.

* Regarding the linkage map, lepmap orders LGs according to the number of markers mapped to it, but this is not the same as the original chromosome orders in the guppy genome. For e.g. chr2 is the largest chromosome, and as such has the most markers, making it "LG1". These were manually changed to match the order of chromosomes in the genome during analyses.

Additional associated data and analysis scripts can be found in the github repository: (archived with Zenodo doi:10.5281/zenodo.5938562)


European Research Council, Award: GuppyCON 758382

Natural Environment Research Council, Award: NE/P013074/1

National Science Foundation, Award: DEB-0623632EF

National Science Foundation, Award: DEB-0808039

National Science Foundation, Award: DEB-1258231

National Science Foundation, Award: DEB-1556884

Biotechnology and Biological Sciences Research Council, Award: BB/K003240/1

Wellcome Trust, Award: WT101650MA