In richelbilderbeek/gcaer: Interface to GCAE

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

In this vignette, we'll use GCAE on the PLINK tutorial's example data, to train and use a GCAE neural network.

Here are the steps:

Setup GCAE
See the genetic data
See the GCAE model
Use GCAE for training

Setup GCAE

First we determine if GCAE is installed and if the PLINK tutorial data is installed.

To determine if GCAE is installed, we'll first load gcaer:

library(gcaer)

Now, gcaer detects if GCAE is installed:

if (!is_gcae_installed()) {
  message("GCAE is not installed")
  message("Tip: use 'gcaer::install_gcae()'")
}

Here we use plinkr to detect if the PLINK tutorial data is installed:

if (!plinkr::is_plink_tutorial_data_installed()) {
  message("PLINK tutorial data is not installed")
  message("Tip: use 'plinkr::install_plink_tutorial_data'")
}

If both requirements are met, we are good to go:

is_gcae_installed <- has_cloned_gcae_repo() # 'is_gcae_installed' is too slow
is_plink_tutorial_installed <- plinkr::is_plink_tutorial_data_installed()
good_to_go <- is_gcae_installed && is_plink_tutorial_installed

Without these requirements met, this vignette will be rather dull :-)

See the genetic data

These are the PLINK tutorial files:

if (good_to_go) {
  plink_tutorial_data_filenames <- plinkr::get_plink_tutorial_data_filenames()
  plink_tutorial_data_filenames
}

We will be using hapmap1 files, that are in the non-binary/plain-text PLINK file format:

if (good_to_go) {
  plain_text_plink_filenames <- stringr::str_subset(
    string = plink_tutorial_data_filenames,
    pattern = "hapmap1"
  )
  plain_text_plink_filenames
}

As GCAE needs these files to be in binary PLINK format, we convert the text files to binary files and store these in a temporary folder:

if (good_to_go) {
  temp_folder <- plinkr::get_plinkr_tempfilename()
  binary_plink_filenames <- plinkr::make_bed(
    base_input_filename = tools::file_path_sans_ext(
      plain_text_plink_filenames[1]
    ),
    base_output_filename = file.path(temp_folder, "hapmap1_binary")
  )
  binary_plink_filenames
}

The three PLINK binary files are the .fam and .bim, .bed files.

This is the .fam file:

n_individuals <- "[unknown]"
if (good_to_go) {
  fam_filename <- stringr::str_subset(
    binary_plink_filenames,
    pattern = "\\.fam$"
  )
  fam_table <- plinkr::read_plink_fam_file(fam_filename)
  n_individuals <- nrow(fam_table)
  knitr::kable(utils::head(fam_table))
}

This table shows us the r n_individuals individuals.

This is the .bim file:

n_snps <- "[unknown]"
if (good_to_go) {
  bim_filename <- stringr::str_subset(
    binary_plink_filenames,
    pattern = "\\.bim$"
  )
  bim_table <- plinkr::read_plink_bim_file(bim_filename)
  n_snps <- nrow(bim_table)
  knitr::kable(utils::head(bim_table))
}

The .bim file shows the information of the r n_snps SNPs.

The .bed file connects the individuals and the SNPs:

if (good_to_go) {
  bed_filename <- stringr::str_subset(
    binary_plink_filenames,
    pattern = "\\.bed$"
  )
  bed_table <- plinkr::read_plink_bed_file(
    bed_filename,
    names_loci = bim_table$id,
    names_ind = fam_table$fam
  )
  testthat::expect_equal(ncol(bed_table), n_individuals)
  testthat::expect_equal(nrow(bed_table), n_snps)
  knitr::kable(utils::head(bed_table[, 1:10]))
}