Litter Diallel

Summary

This package includes functions for reproducing the analysis in our litter diallel manuscript.

Installation

You can install litterDiallel using the following steps. First, please make sure devtools is installed in R.

1. Install MCMCglmm (for compatibility, must be version 2.25) in R:

   devtools::install_version("MCMCglmm", version = "2.25", repos = "http://cran.us.r-project.org")

2. Install litterDiallel (with the vignette, recommended):

   devtools::install_github("mauriziopaul/litterDiallel", build_vignettes=TRUE, build = TRUE, build_opts = c("--no-resave-data", "--no-manual"))

   or

   Install litterDiallel (without the vignette).

   devtools::install_github("mauriziopaul/litterDiallel")

Using the Package

3. You should then be able to load the package in R:

   library(litterDiallel)

4. To load the data set, litters, use:

   data("litters")

5. For an overview of the analysis, see the vignette:

   browseVignettes("litterDiallel")

Analysis Summary

The included data file, litter-diallel.csv, has the following column names, where each of 4,448 observed litters is represented by a single row. All counts represent animals that are observed at the time of weaning (~ 3 weeks after birth):

• Dam_Founder (factor): The inbred mouse strain name of the female parent, where the following abbreviations are used: A/J (AJ), C57BL/6J (B6), 129S1/SvImJ (129S1), NOD/ShiLtJ (NOD), NZO/HlLtJ (NZO), CAST/EiJ (CAST), PWK/PhJ (PWK), and WSB/EiJ (WSB).
• Sire_Founder (factor): The inbred mouse strain name of the male parent.
• PupGeno (factor): The strain-cross name of the F1 offspring, given as their dam-by-sire cross (strainDam x strainSire).
• WeanDate (factor): The date (DD/MM/YY) that animals were weaned into new cages and separated from their parents.
• YearMonth (factor): The date (YearMonth-YYYY-MM) of weaning.
• litterorder (factor): The parity, or litter birth order based on dam, is provided as a factor.
• litternum (integer): The litter number (litterorder - 1) is provided as an integer, where the first litter is 0, and each subsequent litter is numbered starting at 1.
• First_Litter (integer): A binary variable taking the values 0 or 1, indicating whether it is the first litter born to the given dam.
• Males (integer): The number of male pups in the litter.
• Females (integer): The number of female pups in the litter.
• Weaned (integer): The total number of (male and female) pups in the litter.
• Male_Prop (numeric): The proportion of male pups to overall pups in the litter.

The order of the columns in the data set does not matter.

We use the functions diallelMatrixMaker and diallelMatrixMakeAndRotate to generate design matrices for modeling the different classes of effects. After reading in the data, these functions expect: the name of the data frame object, the dam column name, the sire column name, and two random effect (batch, batch.1) column names.

We then use the MCMCglmm function from the MCMCglmm package (version 2.25) to analyze our data by fitting (generalized) linear mixed models, including the overdispersed zero-truncated Poisson, binomial, or Gaussian model.

Notes

• The "fulls" model (with sex-specific effects) and the "a,v" model (with additive and epistatic effects only) can now be run with this package.

• For the model inclusion probability analysis, BayesDiallel and BayesSpike must be installed. The instructions are here.

• For some simple, miscellaneous functions useful for this analysis, install the following package, PLMcctools:

  devtools::install_github("mauriziopaul/PLMcctools")

• To install other suggested packages, use:

  install.packages(c("tools", "data.table", "xtable"))