Marco Antônio Peixoto
Summer/2023

1. Introduction

Cross-validation schemes are a fundamental concept in plant breeding programs that use Genomic selection as a tool to harness genetic gain. One important way to measure how good is your model is the implementation of cross-validation schemes (aka CV schemes). Following the proposition made by Jarquin et al. (2018), they can be broken down into four types:

Here, I put some effort together to combine some scripts that implement those cross-validation schemes.

2. How to Implement it

The first step will be to organize the phenotypic information and create the partitions to the cross-validation scheme (CV). Generally, we can divide the phenotypic dataset into two:

Testing set: Population where the phenotype will be predicted based on the markers and the information from the training set.

Training set: Population with the information that we use to calibrate the model.

Cross-validation scheme 2

• Cross-validation [html]

Cross-validation scheme 1

• Cross-validation [html]

Any questions about the analyses, please, contact me!

Dr. Marco Antonio Peixoto
Email: deamorimpeixotom@ufl.edu
Page: https://marcopxt.github.io/