From 0d8f31c67902d947bc20a161e6abba1ef39d8168 Mon Sep 17 00:00:00 2001
From: Jonathan Kitt <jonathan.kitt@inrae.fr>
Date: Mon, 3 Feb 2025 13:51:44 +0100
Subject: [PATCH] update calculate_lr(): lr + lr stand
---
R/calculate_lr.R | 99 +++++++++++++++++++++++++++++----------------
R/standardise_lr.R | 3 --
man/calculate_lr.Rd | 12 ++++--
3 files changed, 72 insertions(+), 42 deletions(-)
diff --git a/R/calculate_lr.R b/R/calculate_lr.R
index 43d1da1..72007f8 100644
--- a/R/calculate_lr.R
+++ b/R/calculate_lr.R
@@ -1,52 +1,81 @@
-#' Calculate lr values from summary file
+#' Calculate lr values from summary file and standardize lr values
#'
-#' @param summary_file imported summary file (usually named 'summary')
+#' @param axiom_data imported Axiom files (obtained by calling read_axiom())
+#' @param standard whether or not to standardize the lr values ('none', 'mean' - the default , 'ref' or 'both')
+#' @param ref name of the sample to use for standardization (must be defined if method is "ref" or "both", defaults to NULL)
#'
#' @import dplyr
-#' @import tidyr
#'
#' @return A [tibble()].
#' @export
#'
#' @examples
#' \dontrun{
-#' calculate_lr(summary_file)
+#' calculate_lr(axiom_data, standard = "mean", ref = NULL)
#' }
calculate_lr <- function(summary_file) {
# Set NULL variables
- a <- b <- pivot_longer <- probeset_id <- NULL
-
- # Extract A signal from summary file
- message("Extracting A signal from summary file")
- signal_a <- summary_file |>
- dplyr::filter(!grepl(pattern = "NP", x = probeset_id)) |>
- dplyr::filter(grepl("AX-[0-9]*-A", probeset_id)) |>
- dplyr::mutate(probeset_id = stringr::str_remove(string = probeset_id, pattern = "-A")) |>
- tidyr::pivot_longer(cols = -probeset_id,
- names_to = "file_name",
- values_to = "signal_a")
-
- # Extract B signal from summary file
- message("Extracting B signal from summary file")
- signal_b <- summary_file |>
- dplyr::filter(!grepl(pattern = "NP", x = probeset_id)) |>
- dplyr::filter(grepl("AX-[0-9]*-B", probeset_id)) |>
- dplyr::mutate(probeset_id = stringr::str_remove(string = probeset_id, pattern = "-B")) |>
- tidyr::pivot_longer(cols = -probeset_id,
- names_to = "file_name",
- values_to = "signal_b")
-
- # Calculate lr values
- message("Calculating lr values")
- lr <- signal_a |>
- dplyr::left_join(signal_b) |>
- dplyr::mutate(lr = log2(sqrt(signal_a^2 + signal_b^2)),
- .keep = "unused")
-
- # Assign to Global Environment
- assign(x = "lr", value = lr, pos = ".GlobalEnv")
+ # a <- b <- pivot_longer <- probeset_id <- NULL
+
+ # If standard = "none", calculate lr values
+ if (standard == "none") {
+ message("Calculating lr values")
+ lr <- axiom_data |>
+ dplyr::mutate(lr = log2(sqrt(signal_a^2 + signal_b^2)),
+ .after = signal_b)
+ }
+
+ # If standard = "ref" or "both", check if ref is defined
+ if(standard %in% c("ref", "both")){
+ print("Checking if reference sample is defined")
+ if(is.null(ref)){
+ stop("Reference sample is not defined")
+ }
+ if(!is.null(ref) & !ref %in% unique(axiom_data$file_name)) {
+ stop("Reference sample doesn't exist in the list of samples")
+ }
+ }
+
+ # If standard = "mean", calculate lr + lr_st_mean
+ if (standard == "mean") {
+ message("Calculating lr values and lr_st_mean values")
+ lr <- axiom_data |>
+ dplyr::mutate(lr = log2(sqrt(signal_a^2 + signal_b^2)),
+ .after = signal_b) |>
+ dplyr::mutate(lr_mean = mean(lr, na.rm = TRUE),
+ .by = probeset_id) |>
+ dplyr::mutate(lr_st_mean = lr - lr_mean,
+ .after = lr) |>
+ dplyr::select(-lr_mean)
+ }
+
+ # If standard = "ref", calculatelr + lr_st_ref
+ if (standard == "ref") {
+ message("Calculating lr values and lr_st_ref values")
+ lr <- axiom_data |>
+ dplyr::mutate(lr = log2(sqrt(signal_a^2 + signal_b^2)),
+ .after = signal_b) |>
+ dplyr::mutate(lr_st_ref = lr - lr[file_name == ref],
+ .by = probeset_id,
+ .after = lr)
+ }
+
+ # If standard = "both", calculate lr + lr_st_ref & lr_st_mean
+ if (standard == "both") {
+ message("Calculating lr values, lr_st_mean and lr_st_ref values")
+ lr <- axiom_data |>
+ dplyr::mutate(lr = log2(sqrt(signal_a^2 + signal_b^2)),
+ .after = signal_b) |>
+ dplyr::mutate(lr_st_ref = lr - lr[file_name == ref],
+ lr_mean = mean(lr, na.rm = TRUE),
+ .by = probeset_id) |>
+ dplyr::mutate(lr_st_mean = lr - lr_mean) |>
+ dplyr::select(file_name:lr, lr_st_mean, lr_st_ref, genotyping_call)
+ }
+
+ return(lr)
}
diff --git a/R/standardise_lr.R b/R/standardise_lr.R
index f7b3495..040110f 100644
--- a/R/standardise_lr.R
+++ b/R/standardise_lr.R
@@ -56,7 +56,4 @@ standardise_lr <- function(lr, method = "mean", ref = NULL) {
dplyr::select(-lr_mean)
}
- # Assign to Global Environment
- assign(x = "lr", value = lr, pos = ".GlobalEnv")
-
}
diff --git a/man/calculate_lr.Rd b/man/calculate_lr.Rd
index d9a2191..a9cf95d 100644
--- a/man/calculate_lr.Rd
+++ b/man/calculate_lr.Rd
@@ -2,21 +2,25 @@
% Please edit documentation in R/calculate_lr.R
\name{calculate_lr}
\alias{calculate_lr}
-\title{Calculate lr values from summary file}
+\title{Calculate lr values from summary file and standardize lr values}
\usage{
calculate_lr(summary_file)
}
\arguments{
-\item{summary_file}{imported summary file (usually named 'summary')}
+\item{axiom_data}{imported Axiom files (obtained by calling read_axiom())}
+
+\item{standard}{whether or not to standardize the lr values ('none', 'mean' - the default , 'ref' or 'both')}
+
+\item{ref}{name of the sample to use for standardization (must be defined if method is "ref" or "both", defaults to NULL)}
}
\value{
A \code{\link[=tibble]{tibble()}}.
}
\description{
-Calculate lr values from summary file
+Calculate lr values from summary file and standardize lr values
}
\examples{
\dontrun{
-calculate_lr(summary_file)
+calculate_lr(axiom_data, standard = "mean", ref = NULL)
}
}
--
GitLab