In tidymass/massqc: Quality assessment of LC-MS based metabolomics data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warning = FALSE,
  message = TRUE,
  out.width = "100%"
)

Sample plot

library(massdataset)
library(ggplot2)
library(tidyverse)
library(massqc)
data("expression_data")
data("sample_info")
data("variable_info")

object =
  create_mass_dataset(
    expression_data = expression_data,
    sample_info = sample_info,
    variable_info = variable_info
  )

object %>%
  massqc_sample_boxplot()

object %>%
  log(10) %>%
  massqc_sample_boxplot()

object %>%
  log(10) %>%
  massqc_sample_boxplot(color_by = "class")

object %>%
  log(10) %>%
  massqc_sample_boxplot(fill_by = "class") +
  ggsci::scale_fill_lancet()

object %>%
  log(10) %>%
  massqc_sample_boxplot(
    fill_by = "class",
    color_by = "class",
    point = TRUE,
    point_alpha = 0.3
  ) +
  ggsci::scale_fill_lancet()

object %>%
  log(10) %>%
  massqc_sample_boxplot(color_by = "class",
                 point = TRUE,
                 point_alpha = 0.3) +
  ggsci::scale_color_lancet()

PCA plot

object =
  create_mass_dataset(
    expression_data = expression_data,
    sample_info = sample_info,
    variable_info = variable_info
  )

object %>%
  massqc_pca()

object %>%
  massqc_pca(color_by = "class")

object %>%
  scale %>%
  massqc_pca(color_by = "class")

object %>%
  scale %>%
  massqc_pca(color_by = "class", frame = FALSE) +
  ggsci::scale_fill_lancet()

object %>%
  scale %>%
  massqc_pca(color_by = "class", frame = FALSE) +
  ggsci::scale_fill_lancet() +
  ggrepel::geom_text_repel(aes(label = sample_id))

object %>%
  scale %>%
  massqc_pca(color_by = "class", frame = FALSE) +
  ggsci::scale_fill_lancet() +
  ggrepel::geom_text_repel(aes(label = ifelse(class == "QC", sample_id, NA)))


object %>%
  massqc_pca_pc1()

object %>%
  massqc_pca_pc1(color_by = "class")

object %>%
  scale %>%
  massqc_pca_pc1(color_by = "class")

object %>%
  scale %>%
  massqc_pca_pc1(
    color_by = "class",
    order_by = "injection.order",
    point_alpha = 1,
    point_size = 5
  ) +
  ggsci::scale_color_lancet()

object %>%
  scale %>%
  massqc_pca_pc1(
    color_by = "class",
    order_by = "injection.order",
    point_alpha = 1,
    point_size = 5, 
    desc = TRUE
  ) +
  ggsci::scale_color_lancet()

Missing values

object =
  create_mass_dataset(
    expression_data = expression_data,
    sample_info = sample_info,
    variable_info = variable_info
  )

##show missing values plot
show_missing_values(object)

show_missing_values(object[1:10,], cell_color = "white")

###only show features with mz < 100
object %>%
  activate_mass_dataset(what = "variable_info") %>%
  dplyr::filter(mz < 100) %>%
  show_missing_values(cell_color = "white",
                      show_row_names = TRUE,
                      row_names_side = "left")

show_sample_missing_values(object, color_by = "class", order_by = "na")

show_variable_missing_values(object, color_by = "rt") +
  scale_color_gradient(low = "skyblue", high = "red")

RSD

library(massdataset)
library(ggplot2)
library(tidyverse)
data("expression_data")
data("sample_info")
data("variable_info")

object =
  create_mass_dataset(
    expression_data = expression_data,
    sample_info = sample_info,
    variable_info = variable_info
  )

object %>% 
  massqc_rsd_plot()

object %>% 
  massqc_rsd_plot(color_by = "rsd")

object %>% 
  massqc_rsd_plot(color_by = "rsd", order_by = "rsd")

object %>% 
  massqc_rsd_plot(color_by = "rsd", point_alpha = 1) +
  scale_color_gradient(low = "skyblue", high = "red") +
  geom_hline(yintercept = 0.3, color = "red")

object %>% 
  activate_mass_dataset(what = "sample_info") %>% 
  filter(class == "Subject") %>% 
  massqc_rsd_plot(color_by = "rsd", point_alpha = 1) +
  scale_color_gradient(low = "skyblue", high = "red") +
  geom_hline(yintercept = 0.3, color = "red")