This reproducible R Markdown
analysis was created with workflowr (version
1.7.2). The Checks tab describes the reproducibility checks
that were applied when the results were created. The Past
versions tab lists the development history.
The R Markdown file has unstaged changes. To know which version of
the R Markdown file created these results, you’ll want to first commit
it to the Git repo. If you’re still working on the analysis, you can
ignore this warning. When you’re finished, you can run
wflow_publish to commit the R Markdown file and build the
HTML.
Great job! The global environment was empty. Objects defined in the
global environment can affect the analysis in your R Markdown file in
unknown ways. For reproduciblity it’s best to always run the code in an
empty environment.
The command set.seed(20250829) was run prior to running
the code in the R Markdown file. Setting a seed ensures that any results
that rely on randomness, e.g. subsampling or permutations, are
reproducible.
Great! You are using Git for version control. Tracking code development
and connecting the code version to the results is critical for
reproducibility.
The results in this page were generated with repository version
bb0b635.
See the Past versions tab to see a history of the changes made
to the R Markdown and HTML files.
Note that you need to be careful to ensure that all relevant files for
the analysis have been committed to Git prior to generating the results
(you can use wflow_publish or
wflow_git_commit). workflowr only checks the R Markdown
file, but you know if there are other scripts or data files that it
depends on. Below is the status of the Git repository when the results
were generated:
Note that any generated files, e.g. HTML, png, CSS, etc., are not
included in this status report because it is ok for generated content to
have uncommitted changes.
These are the previous versions of the repository in which changes were
made to the R Markdown (analysis/about.Rmd) and HTML
(docs/about.html) files. If you’ve configured a remote Git
repository (see ?wflow_git_remote), click on the hyperlinks
in the table below to view the files as they were in that past version.
This repository accompanies the manuscript “Kernel-Based
Integration of Near-Infrared Spectroscopy and Genomic Data for Enhanced
Predictive Performance” and documents the data-processing,
kernel construction, prediction, and visualization workflows used in the
study.
The project evaluates whether kernel-based integration of
genomic markers, near-infrared spectroscopy
(NIRS), and weather covariates can improve
prediction for agronomic traits in multi-environment maize trials. The
analyses were developed under a hierarchical Bayesian framework in
R, with reproducible workflows organized through
workflowr.
Study overview
The study was based on multi-environment field trials conducted in
College Station, Texas, during 2011
and 2012, under two contrasting water management
regimes: well-watered (WW) and water-stressed
(WS). These combinations generated four macro-environments used
throughout the prediction analyses:
CS11_WS
CS11_WW
CS12_WS
CS12_WW
The repository integrates three main sources of information:
Genomic data: genome-wide SNP markers obtained by
genotyping-by-sequencing.
Phenomic data: whole-kernel NIRS spectra measured
after harvest.
Weather data: daily meteorological covariates
summarized and structured for environmental kernel construction.
The target traits analyzed in the manuscript are:
Grain yield (GY)
500-kernel weight (KW)
Analytical scope
The repository contains the workflows required to:
harmonize phenotypic, genomic, spectral, and environmental
data;
construct kernel matrices for genomic (G),
phenomic/NIRS (P), environmental (W),
and interaction structures;
fit single- and multi-kernel prediction models using
BGLR;
compare Gaussian and arc-cosine
kernels;
evaluate predictive performance under four validation schemes:
CV1: untested genotypes;
CV2: untested genotypes with partial
information;
CV0: untested environment;
CV00: untested genotypes and untested
environments;
summarize variance components and predictive ability across model
classes.
Main findings
The central hypothesis of the study was that integrating genomic and
NIRS information through kernel methods could improve predictive ability
relative to single-source models. However, the current analyses indicate
a more restrictive result:
genomic prediction (G) was the most robust
information source across validation scenarios;
phenomic prediction based on NIRS (P) showed lower
and less stable predictive performance;
the environment effect (E) explained a substantial
portion of the observed variation;
adding weather covariates (W) or explicitly
modeling interaction kernels did not consistently improve
prediction;
integrating G + P did not produce
clear gains in predictive ability under the evaluated conditions.
These results suggest that, in this dataset, NIRS captured a strong
environment-dependent physiological signal, but did not provide a
sufficiently stable complementary signal to improve genomic prediction
across environments.
Repository structure
The project is organized into the following directories:
analysis/: main R Markdown workflows,
manuscript-oriented analyses, and website source files;
code/: auxiliary R functions and
project utilities;
data/: input datasets used in the
analyses;
output/: derived matrices,
intermediate objects, and prediction summaries;
docs/: rendered website files for
GitHub Pages.
Reproducibility
All analyses were implemented in R and are intended
to be reproducible through the project environment. Package dependencies
can be restored with renv, and the website/reporting
workflow is managed with workflowr.
A typical local setup is:
install.packages("renv")
renv::restore()
Citation and contact
If you use this repository, please cite the corresponding manuscript
when available.
Author / Contact
Weverton Gomes da Costa
Departamento de Estatística, Universidade Federal de Viçosa (UFV)
Email: weverton.costa@ufv.br