In MassDynamics/lfq_processing: Processes Proteomics Files for Quantification.

knitr::opts_chunk$set(fig.cap = NULL, fig.path = params$output_figure)

library(data.table)
library(ggplot2)
library(ggrepel)
library(GGally)
library(umap)
library(FactoMineR)
library(factoextra)
library(corrplot)
library(viridis)
library(ggpubr)
library(Hmisc)
library(plotly)
library(stringr)
library(bit64)

num_files <- expdes[, .N]
run_per_condition <- expdes[, .(countRepMax = .N), by = .(experiment)]
setnames(run_per_condition, "experiment", "condition")

# for fractions, create file name from mqExperiment and Fraction
if (!("file_name" %in% colnames(expdes))){
  expdes$file_name = paste(expdes$experiment, " - ", expdes$Replicate)
}

# pep/mod pep
mod_pept_int_rep <- merge(
  run_per_condition,
  mod_pept_int[Imputed == 0, .(Repcount = .N), by = .(id, condition)],
  by = c("condition")
)
mod_pept_int_rep[, repPC := Repcount/countRepMax]
mod_pept_id_in_a_cond <- mod_pept_int_rep[repPC >= 0.5, unique(id)]
mod_pept_int[, Valid := 0L]
mod_pept_int[id %in% mod_pept_id_in_a_cond, Valid := 1L]
rm(mod_pept_id_in_a_cond, mod_pept_int_rep)

pept_int_rep <- merge(
  run_per_condition,
  pept_int[Imputed == 0, .(Repcount = .N), by = .(id, condition)],
  by = c("condition")
)
pept_int_rep[, repPC := Repcount/countRepMax]
pept_id_in_a_cond <- pept_int_rep[repPC >= 0.5, unique(id)]
pept_int[, Valid := 0L]
pept_int[id %in% pept_id_in_a_cond, Valid := 1L]
rm(pept_id_in_a_cond, pept_int_rep)

# proteins  
prot_int_rep <- merge(
  run_per_condition,
  prot_int[Imputed == 0, .(Repcount = .N), by = .(id, condition)],
  by = c("condition")
)
prot_int_rep[, repPC := Repcount/countRepMax]
prot_id_in_a_cond <- prot_int_rep[repPC >= 0.5, unique(id)]
prot_int[, Valid := 0L]
prot_int[id %in% prot_id_in_a_cond, Valid := 1L]
rm(prot_id_in_a_cond, prot_int_rep)

cvdt <- mod_pept_int[Imputed == 0][, countRep := .N, by = .(id, condition)]
cvdt[, countRepMax := max(countRep), by = .(id, condition)]
cvdt[, ReplicatePC := countRep/countRepMax]
cvdt[, intensity := as.double(intensity)]
cvdt <- cvdt[ReplicatePC >= 0.5, .(cv = sd(intensity)/mean(intensity)), by = .(id, condition)]

p <- ggplot(cvdt, aes(x=cv, fill=condition, colour=condition)) +
  geom_density(alpha=0.4) +
  theme_minimal() +
  scale_x_continuous("% CV", labels = scales::percent) +
  ggtitle("Modified Peptides - LFQ intensity CV")

ggplotly(p) %>% config(displayModeBar = T, 
                        modeBarButtons = list(list('toImage')),
                        displaylogo = F)