Pedixplorer: a BioConductor package to create, filter and draw pedigree

This is the new version of kinship2 package. Initially a set of functions to view pedigrees while developing models that use kinship matrices, the functions were useful enough to put into a package of its own. It has now an S4 class for pedigrees, a function to computes the kinship matrix from a Pedigree object, and pedigree plotting routines that adhere to many of the standards for genetics counselors.

Try today the Pedixplorer shiny app to easily use the package.

Installation

With bioconda

mamba create -n env_pedixplorer bioconda::bioconductor-pedixplorer
mamba activate env_pedixplorer

In R from Github

if (!require("remotes", quietly = TRUE))
    install.packages("remotes")

remotes::install_github("louislenezet/Pedixplorer",
    build_vignettes=TRUE
)

In R from Bioconductor

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("Pedixplorer")

Main functions

The Pedigree() Function

Pedigree() is a function that creates an S4 class Pedigree object. The core slot of the Pedigree object is the ped slot built from having a row per person, linked by the father id and mother id. Other relationships can be specified, and affection status can be a matrix of multiple categories in the rel slot. All the informations about how the affection and availability have to be draw are stored respectively in scales$fill and scales$border slots They are used to fill and color the border for each elements of the Pedigree graph.

The generate_colors() Function

generate_colors() is a function that generates a color palette for an affection status. This function is used by the Pedigree() function to generate the scales$fill and scales$border slots. The user can also use this function to generate a color palette for a specific affection status that will be added to the Pedigree object.

The plot() Method

A Pedigreee plot() S4 method is available to plot the object as a "family tree", with relatives of the same generation on the same row, and affection statuses divided over the plot symbol for each person. This function is designed in two steps:

1. First the Pedigree object is converted into a data frame with all the elements needed to plot the Pedigree (i.e. boxes, lines, text, etc.). This is done by the ped_to_plotdf() function.
2. Then the data frame is plotted using the plot_fromdf() function.

The kinship() Funtion

kinship() is a function that creates the kinship matrix from a Pedigree object. It is coded for dyplotype organisms, handling all relationships that can be specified for the Pedigree object, including inbreeding, monozygotic twins, etc. A recent addition is handling the kinship matrix for the X and Y chromosomes.

The ped_shiny() Function

To help anyone to easily use all the main functions of the package a shiny app has been created, allowing you to import your data, normalise it, select the family and filter the resulting `Pedigree` object before visualising it. You'll also be able to download the resulting data and plot. The application is also available on a Virtual Machine accessible at pedixplorer.univ-rennes.

Other Notable Functions

• useful_ind() automatically find the individuals close to a given set of individuals, allowing to split the Pedigree in smaller family for an easier representation.

• shrink() will shrink a Pedigree to a given size, keeping the most informative individuals for a single affection variable.

• fix_parents() will add parents for children who have a mother or dad listed that is not already included. It will also fix the sex status for the parent if it is mis-specified. This is useful to use before creating the Pedigree object.

Example

Here is a simple example that show how to represent a complex pedigree with a lot of different information.

library(Pedixplorer)
library(dplyr)
data("sampleped")
data("relped")

# Create the Pedigree object
pedi <- Pedigree(sampleped, relped, missid = NA) %>%
    generate_colors( # Add a new affection information
        col_aff = "num", is_num = TRUE,
        keep_full_scale = TRUE, breaks = 2,
        threshold = 3,
        colors_aff = c("#8B7355", "#FFA500"),
        colors_unaff = c("#8aca25", "#3fb7db")
    ) %>%
    is_informative( # Set which individuals are informative
        col_aff = "num", informative = "AvAf"
    ) %>%
    useful_inds(
        keep_infos = TRUE, # Keep available or affected parents 
        max_dist = 2 # Maximum distance from informative individuals
    )

proband(ped(pedi)) <- isinf(ped(pedi)) # Set informative individuals as proband

png("MyPedigree.png", width = 1000, height = 600)
plot_list <- plot(
    pedi,
    symbolsize = 1.5, # Increase the symbole size
    title = "My pedigree", # Add a title
    legend = TRUE, # Add the legend
    leg_symbolsize = 0.02, # Set the symbole size of the legend
    leg_loc = c(0.5, 0.9, 0.8, 1.1), # Specify the legend location
    lwd = 0.5, # Set the line width
    ggplot_gen = TRUE, # Use ggplot2 to draw the Pedigree
    tips = c(
        "id", "avail",
        "affection",
        "num", "dateofbirth"
    ) # Add some information in the tooltip
)
dev.off()

# Plot the Pedigree with plotly to have an interactive plot
plotly::ggplotly(
    plot_list$ggplot,
    tooltip = "text"
) %>%
    plotly::layout(hoverlabel = list(bgcolor = "darkgrey"))

View Interactive Pedigree

Documentation and News

To view documentation start R and enter:

library(Pedixplorer)
help(package="Pedixplorer")

# Or to view the vignettes
browseVignettes("Pedixplorer")

# Or to see the news
utils::news(package="Pedixplorer")