Skip to content

Latest commit

 

History

History
62 lines (43 loc) · 4.38 KB

README.md

File metadata and controls

62 lines (43 loc) · 4.38 KB
Raw

Code and example data for:

Stoffel, MA, Johnston, SE, Pilkington, JG, Pemberton, JM: Genetic architecture and lifetime dynamics of inbreeding depression in a wild mammal. Nature Communications 12, 2972 (2021). https://doi.org/10.1038/s41467-021-23222-9.



Project

Across the animal and plant kingdoms, matings between relatives often result in reduced fitness of their offspring, a phenomenon termed inbreeding depression. This project aimed at gaining a better understanding of the genetic basis of inbreeding depression by analysing genomic and survival data from a long-term study of wild Soay sheep on the remote Scottish St. Kilda archipelago. At the heart of the analyses are runs of homozygosity, a signature feature of inbred genomes, which we used to pinpoint genetic regions containing deleterious mutations.

Runs of homozygosity among individuals, length classes and across the genome .

Code structure

This repository contains the analysis code for our paper, in order 1-7.

Script 1-3 process data, specifically:
1_preprocessing_and_ROH_calling: Preprocesses genotype data and calls ROH
2_calculate_fitness_parameters_v2: Transform tables from the Soay sheep database into a table with annual fitness measures
3_combine_ROH_and_fitness_data: Calculates FROH and combines fitness and FROH data

Scripts 4-7b contain the main analyses in the paper:
4_ROH_patterns: First part of the paper, including Figure 1
4b_ROH_recombination Running window analyses / plots for ROH density and recombination 5a_survival_models: INLA animal models to quantify inbreeding depression in survival
5b_survival_models_figure: Creates Figure 3 based on INLA output
6_alt_gwas_annual_survival_bothA_sep: ROH GWAS, needs to run on a cluster
7a_gwas_postprocessing_bothA_sep: Takes GWAS results, checks for errors and makes a Manhattan plot
7b_gwas_top_snps_sep: Visualises genetic diversity and GWAS estimates in the genomic vicinity of GWAS peaks
99_add_suppl_figures: Creates some additional figures for Supplementary Material
99_make_pcs_for_gwas: Get PCs in the right format for use in GWAS.

Data and reproducibility

All analysis scripts (4-7b) can be run with data provided in the folders example_data and example_output. For a smooth experience with running these scripts, download the complete repository and change the names of example_data and example_output to data and output, respectively. Examples are based on a random subset of 100 individuals, so results will be different from those reported in the paper.

The full data can be downloaded from Zenodo: DOI

Versions and dependencies

Running the code depends on a series of R packages, which mostly come from CRAN, but some are development versions on GitHub or stored on private websites. The versions of all dependencies are stored alongside the code in the renv.lock file. You can either install these packages yourself while running the code, or you can use the renv package to setup everything for you. To do so, download or clone this repository somewhere onto your computer. Then install renv with:

if (!requireNamespace("remotes"))
  install.packages("remotes")

remotes::install_github("rstudio/renv")

You can then simply run renv::init() from the directory. This will find the renv.lock file and create an renv folder. The folder will contain a private R library with all the packages used in these scripts. This might take a while as there are a lot of packages to download and install. More info here: https://rstudio.github.io/renv/. Two packages have to be downloaded from private websites (INLA and AnimalINLA); the websites links are given in the scripts where they are used.

SNP chip preprocessing and imputation.

The scripts for some of the data preprocessing and for the genotype imputation are stored in the following repositories:

  1. SNP chip merging and LD decay
  2. Genotype imputation and cross-validation
  3. Imputation output postprocessing