Cowpea adaptation to high temperatures in Africa

Detect candidate genes for adaptation to high temperature in African cowpea using via environmental genome wide association study (envGWAS).

• The current manuscript is available as:
  Akakpo R, Lee EJ, Pacheco JB, Rios EF, Kantar MB, Ousmane B, Volz KM, Akinmade H, Getino L, Boote KJ, et al. 2025. Identification of genetic variation associated with high-temperature tolerance in cowpea. bioRxiv. doi:10.1101/2025.04.09.647789

UCR collaborators cowpea GWAS

cowpea_gwas

Description

Using SNP data available from IITA Core accessions, structure of population analysis in ~600 accessions of African cowpea will be performed. Various population Genetic analysis including population-based estimate of recombination rate Ped/pop method. Using the same data we will carrie out environmental genome wide association study .

Genotyping datasets

• Genotyping dataset uses UCR Minicore Cluster file generated by Muñoz-Amatriaín M SNP genotyping Matrix

SNP position

• SNP positions from the iSelect design file are listed Github SNP_Utils

Passport data

• A Full passport data includes market collections & breeding material was provided by IITA (~1400 accessions) Full passport data
• Among them, some samples have localities, but no latitude and longitude. We curate the data for a cleaned passport data file that does not include market collection or breeding material. Cheking for matching information between country, Latitude/Longitude and location information will be done. Maybe possible to supplement with inferred latitude & longitude.
• Cleaned passport data will be upload on GitHub Cleaned passport data

Bioclimatic data source

• WorldClim is environmental data worldwide. We will focus on mean precipitation and temperature data, but also on growing month data WorldClim

Analysis Steps and Tools

EnvGWAS

1. Performing envGWAS for BIO 1-19
2. Performing envGWAS for Monthly Rainfall, Max temperature and Min temperature

Population structure and differentiation

1. Run structure analysis using STRUCTURE. Try the analysis for correlated & uncorrelated allele frequency i.e. models with admixture & no admixture As structure is time consuming for high dataset, get use LD pruning in Plink to get a subset of the total SNPs to run structure analysis. Introduce iterations in the parameters file, important for running structure

2. Run Procrustes analysis of PCA [Procruste PCA](PLoS Genet. 8: e1002886)

3. Fst Population differentiation test

Estimate FST for each SNP between subgroups (based on structure subgroups and/or geographic groups) and Identify outliers SNPs candidate for adaptive selection.

Gene association tests

1. Identify the gene underling candidates SNPs using Bedtools intersect Bedtools, Snp2Gene Python program and appropriate GFF annotation files

2. Perform Gene Ontology Tests to identify gene functions using the standard Fisher’s exact tests implemented in the R package TopGO

3. LD of "hits" from iSelect to resequencing can use 36 cowpea SNP calls

Maps of collection localities

• Collection locality & temperature Maps

Repository Organization

00_format

The directory contains scripts for formating data.

1. Extract_Monthly_WorlClim_data.R is extracts mean monthly data from raster wordclim data;
2. Extract_Growing_Period_Clim_data.R is used formats genotype and phenotype text file into plink format (.tped, .tfam);
3. ped2binary.sh converts plink text files into binary and produces set of plink files for envgwas

01_envGWAS

This directory contains scripts to run envGWAS.

1. rel_mtx.sh calculates centered relatedness matrices for envgwas
2. mlm.sh is used to fit a mixed-linear models with gemma
3. prep_gemma_output_for_manathan_plot.R is an R script used to transform gemma output files into a suitable files for manhattan plot
4. manhattan_plot.R is an R script used to produce manhattan plot