The Effect of Haplotype Size on Genomic Selection Accuracy and Epistasis: An Empirical Study in Rice - Dataset
Description
The Effect of Haplotype Size on Genomic Selection Accuracy and Epistasis: An Empirical Study in Rice Author: Maria Montiel Description: This study investigates how haplotype block size impacts genomic selection (GS) accuracy and the detection of epistatic interactions in rice breeding populations. It includes phenotypic and genotypic datasets across three years and uses various statistical models and genomic prediction pipelines. This repository contains all the data and scripts used in the study, including phenotypic and genotypic datasets from multiple rice populations evaluated over three years, alongside R scripts for spatial adjustments, genomic selection analysis, and epistasis modeling. Folders: - 001 Phenotypic Data: Contains raw phenotypic data from MP6-8 (multi-parent) and MP2 (bi-parental) populations (2020–2022 trials). Includes Preliminary Yield Trials (PYT) data. Traits: grain yield, days to heading, chalk, and whole milling. Example file: `project_20mp2.csv` - 002 Genotypic Data: Contains genotypic data in dosage matrix format (0, 1, 2) for MP2, MP6-8, MP4 populations. Includes genotype data for PYTs (2020–2022), both imputed and non-imputed. Also contains "LDdistance.csv" with LSU500 SNP distances. - 003 Scripts: R scripts used for data processing, spatial adjustment, genomic selection, and matrix computation. - README.txt Data Format: - Phenotypic Data: .csv - Genotypic Data: .csv - Scripts: .R For more information, please refer to the "README.txt" file.
Files
Steps to reproduce
This repository contains data and scripts for the study titled *“The Effect of Haplotype Size on Genomic Selection Accuracy and Epistasis in Rice.”* The goal of this study is to explore how haplotype block size impacts genomic selection (GS) accuracy and epistatic interactions in rice breeding populations. Below are the steps to reproduce the genomic prediction analysis. 1. Set Up the Environment - Install R (version 4.0 or higher). - Install necessary R packages: ```r install.packages(c("sommer", "ggplot2", "SpATS")) ``` 2. Prepare the Data - The phenotypic and genotypic data are stored in: - `001 Phenotypic Data/`: Phenotypic data from MP2, MP6-8, and PYT. - `002 Genotypic Data/`: Genotypic data in dosage matrix format (0, 1, 2) for each population. - Example files: `project_20mp2.csv`, `Genotype_raw_MP2.csv`. 3. Data Processing - Use `002-a Spatial adjustment and BLUES MP.R` for spatial adjustment and BLUES calculation for MP materials, and `002-b Spatial adjustment and BLUES PYT.R` for PYT materials. This corrects phenotypic data for spatial variability. 4. Calculate Genetic Relationship Matrices - Run `001 A mat and E mat.R` to calculate additive (A) and epistatic (E) genomic relationship matrices for MP2 and MP6-8 populations. These matrices are stored in `.rds` files. 5. Estimate Heritability and Variance Components - Use `003 Heritability and Variance Components by Trial.R` to estimate narrow-sense heritability (h²) and variance components using additive (A) and additive + epistatic (A+I) models. 6. Perform K-Fold Cross-Validation - Run `004 K-Fold Cross Validation with A and A+I Models.R` to perform K-fold cross-validation to assess genomic prediction accuracy using additive and additive + epistatic models. 7. Genomic Prediction - Use `006 Predictions MP and PYT Genomic Selection.R` to perform genomic selection using MP2 and MP6-8 training sets. The script fits a mixed model and extracts genomic breeding values (GEBVs) to evaluate prediction accuracy. 8. Calculate Linkage Disequilibrium (LD) - Run `007 LD Calculation.R` to calculate and plot linkage disequilibrium (LD) among markers. 9. Output and Results - Outputs include adjusted phenotypic data (e.g., BLUES), genomic relationship matrices, heritability estimates, prediction accuracies, and LD calculations. All results are saved in `.csv` and `.rds` files. By following these steps, you can reproduce the analysis and genomic predictions used in this study.
Institutions
- Louisiana State University