Home

Welcome to the PeeDeeCottonBreedingProgram_GSOutlook wiki!

Here, I'll cover the main steps in running our analysis.

0 Process input data

Arrange the data for input into other programs.

1 Get Overall and Environment means with SAS

Calculate least-squares means for each of the 28 traits analyzed in 14 environments + overall means. Model is fitted to include GxE - environment means are the interaction between genotype and environment.

2 Calculate heritability

Calculate Heritability using the SAS code from Holland and Nyquist (2004). These numbers are used to rescale the prediction abilities so they are more easily comparable between traits.

3 Get GWAS hits

Significant SNPs are discovered using a 20% FDR threshold from a linear model (5 PCs + SNPs) in PLINK1.9. Filtering is done to include SNPs that are found in 5+ resamples of the model.

4 Get base GS model

A basic GBLUP-like model is fitted using RKHS in GBLR.

5 Get GS + GWAS model

Significant SNPs are added to the GBLUP model, and we check for model improvement.

6 Get figures

The figures in the paper are rendered using the data from the other parts of the project.