Visualizing Relatedness Matrices with ggRelatednessMatrix()"
In ggpedigree: Visualizing Pedigrees with 'ggplot2' and 'plotly'

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 3.5,
  fig.height = 2.5,
  echo = TRUE
)

suppressPackageStartupMessages({
  library(BGmisc)
  library(ggplot2)
  library(dplyr)
  library(reshape2)
  library(tidyverse)
})

Introduction

This vignette demonstrates how to calculate and visualize relatedness matrices for animal pedigrees using functions from the BGmisc and ggpedigree packages. We focus on Kluane Red Squirrel Project data, but the workflow generalizes to any pedigree dataset. The ggRelatednessMatrix function is a new addition to the ggpedigree package, which provides a flexible and visually appealing way to explore patterns of genetic and mitochondrial relatedness within families. This function is designed to work seamlessly with the ped2add and ped2mit functions from the BGmisc package, which calculate additive genetic and mitochondrial relatedness matrices, respectively.

We will use the new ggRelatednessMatrix function to create ggplot2-based heatmaps of relatedness matrices, inspired by the approaches in the original squirrel population analyses.

Data Preparation

First, load pedigree data and required libraries. The example data is the life‐history and reproductive metrics for 7,799 individual red squirrels from the Kluane Red Squirrel Project (1987–present). See the redsquirrels data documentation for more details.

library(ggpedigree)
# Load the example data
data("redsquirrels")

Calculating Relatedness Matrices

Here we show how to compute additive genetic and mitochondrial relatedness matrices for a target family, as in the red squirrel analysis.

# sumped <- summarizePedigrees(redsquirrels,
#  famID = "famID",
#  personID = "personID",
#  nbiggest = 5
# )


# Set target family for visualization
fam_filter <- 160 # sumped$biggest_families$famID[3]

# Filter for the largest family, recode sex if needed
ped_filtered <- redsquirrels %>%
  recodeSex(code_female = "F") %>%
  filter(famID == fam_filter)

# Calculate relatedness matrices
add_mat <- ped2add(ped_filtered, isChild_method = "partialparent", sparse = FALSE)
mit_mat <- ped2mit(ped_filtered, isChild_method = "partialparent", sparse = FALSE)

Visualizing Relatedness Matrices

The ggRelatednessMatrix function plots a relatedness matrix as a heatmap, with options for clustering and color customization. Here we show how to create a static plot of the additive genetic relatedness matrix.

p_add <- ggRelatednessMatrix(
  add_mat,
  interactive = FALSE,
  config = list(
    color_palette = c("white", "orange", "red"),
    scale_midpoint = 0.55,
    cluster = TRUE,
    title = "Additive Genetic Relatedness",
    include_upper_triangle = FALSE,
    include_lower_triangle = TRUE
  )
)

p_add

Plot Interactively

Here we plot the mitochondrial relatedness matrix with a custom color palette and interactive features.

p_mit <- ggRelatednessMatrix(
  mit_mat,
  interactive = TRUE,
  config = list(
    color_palette = c("white", "skyblue", "darkblue"),
    scale_midpoint = 0.55,
    cluster = TRUE,
    title = "Mitochondrial Relatedness",
    text_size = 6,
    return_widget = TRUE
  )
)

Dragging the mouse over the plot will show the relatedness value for each pair of individuals. You can also zoom in and out using the mouse wheel or by clicking and dragging.

p_mit

# reduce file size for CRAN
if (interactive()) {
  plotly::partial_bundle(p_mit)
} else {
  plotly::partial_bundle(p_mit, local = TRUE)
}

Customization Options

You can adjust clustering, color scales, labels, and more using the config list. For example, to turn off clustering:

p_add_noclust <- ggRelatednessMatrix(
  add_mat,
  config = list(
    cluster = FALSE, title = "Additive Relatedness (No Clustering)" # ,
    #  geom = "geom_raster"
  )
)
p_add_noclust

Comparison to Base R and Other Approaches

For comparison, here is how the same matrix would be plotted in corrplot, as in earlier squirrel analyses:

if (requireNamespace("corrplot", quietly = TRUE)) {
  corrplot::corrplot(
    as.matrix(add_mat),
    method = "color",
    type = "lower",
    col.lim = c(0, 1.25),
    is.corr = FALSE,
    title = "Additive Relatedness",
    order = "hclust",
    col = corrplot::COL1("Reds", 100),
    tl.pos = "l", tl.col = "black", tl.srt = 5, tl.cex = 0.2,
    mar = c(0, 0, 2, 0)
  )
}

Conclusions

The ggRelatednessMatrix function provides a flexible and visually appealing way to explore patterns of genetic and mitochondrial relatedness within families. This workflow integrates seamlessly with tools from BGmisc and ggpedigree, supporting reproducible, publication-quality visualization for quantitative genetic analysis.