breedersel

The goal of breedersel is to provide a user-friendly interface to do sample selection from a dataset. This package contains one function which runs a Shiny App.

Installation

You can install the development version of breedersel from GitHub with:

# install.packages("pak")
pak::pak("chabrault/breedersel")
# or
devtools::install_github("chabrault/breedersel")

Input dataset

You need a file with a column identifying the genotype and several numerical traits in other columns. File format includes “csv”, “tsv”, “xlsx”, “xls” (with tab selection), “rds”, “txt”,…. It is preferable to have genotype-adjusted values instead of have multiple replicated values for each genotype.

Launch the ShinyApp

You can launch the application by running:

library(breedersel)
breedersel::run_app()

Analysis steps

1. Load input dataset. Modify or delete the columns in the import panel, check the type of columns (numeric or character). Validate which column corresponds to the genotype.

2. View the dataset (optional)

3. Filter the list of genotype by the value of the columns (optional) Select the columns to filter on. Move the slider or select the categories for character columns. You can track the number of rows left in your dataset and add check genotypes. Once you’re done with the filtering, validate the table.

4. Apply a multi-trait selection index (MGIDI) Fill the selection index table:

• Select the trait (double click and select a trait from the list).
• Select min/max/opti for the direction of selection (for example, you may want to maximize yield - so select “max” for yield trait), “opti” corresponds to an optimum value.
• If you have selected “opti”, indicate the optimal value.
• Indicate a numeric relative weight to apply for all the traits (optional, assumed an equal weight for all the traits if not filled).

Once the selection index table is filled, select the intensity of selection (% of genotypes retained), and click on the “Analyze” button. The selection index will be applied on the filtered dataset if this step was not skipped.

5. Custom graphics Drag and drop the columns into the different elements to build a custom plot. Select the type of plot (depends on the input), modify the legend, label, color palette, plot theme, and output the figure.