Script to process HPLC chromatogram data from Chromeleon csv files

The script reads many csv files in one dataframe, assigns the file names as a variable, and offers two versions for plotting. Example files can be found here (in the HPLC folder): https://www.dropbox.com/sh/zzf7y3ijwkat55e/AABUvp7BAARIdYBqZWgk1E37a?dl=0

Put the csv files in your working dir and execute the script

# Angelov 2017
# Script to process HPLC chromatogram data from Chromeleon csv files
# The script reads many csv files in one dataframe, assigns the file names as a variable, and offers two versions for plotting
# HPLC - Dionex
# Chromeleon 7.2.3.0, locale - German, therefore the decimal mark problem.
# 
# example files can be found here (in the HPLC folder):
# https://www.dropbox.com/sh/zzf7y3ijwkat55e/AABUvp7BAARIdYBqZWgk1E37a?dl=0
# put the files in your working dir and execute the script (press Source button on top)
# load packages and define read function
library(purrr)
library(tidyverse)
library(plotly)
library(manipulate)
#this function reads the files and adds a variable with the file name
read_chrom <- function(x) {
  read_delim(x, "\t", col_names = FALSE, skip = 43, locale = locale(decimal_mark = ",")) %>% 
    select(-X2) %>%
    mutate(filename = x) # here some work needed to simplify naming
    }
# read files
files <- list.files(pattern = ".txt")
df <- map(files, read_chrom) %>%  
          map_df(bind_rows) %>% 
          rename(time=X1, sig = X3)
# use samples.tab file to assign sample names, file has two tab-sep, named columns (filename and sample)
# then join with df
# prepare sample template file
write_delim(
  tibble(filenames = files, sample = NA),
  "samples-template.tab",
  delim = "\t") 
# now fill-in the sample names and save the file as "samples.tab"
readline(prompt = "Fill-in sample names in \"samples-template.tab\", save as \"samples.tab\" and hit [enter] to continue")

[1] ""

sampledf <- read_delim("samples.tab", delim = "\t", trim_ws = TRUE)
df2 <- left_join(df, sampledf, by = "filename")
# a shifted plot in plotly, to estimate axis scaling
# if you get an error, install the devel version of ggplot and update plotly
# I introduce a shift using the sample names as factors
# plotly
offset <- 20 # play with y-offset
#df2$sample <- as.factor(df2$sample)  # convert to factor to use in the next step
p <- df2 %>% mutate(yshifted = as.numeric(as.factor(sample)) * offset + sig) %>% # use the factor level to introduce shift.
#                    xshifted = as.numeric(sample) + time) %>%                   # could be used for x-axis shift also.
  plot_ly(x = ~time) %>% 
  add_lines(y = ~ yshifted, color =  ~sample, colors = "Paired")
p

# a ggplot version for saving, using manual adjusting axis scales
  df2 %>% mutate(yshifted = as.numeric(as.factor(sample)) * offset + sig) %>%
  ggplot(aes(x = time, y = yshifted)) + 
  geom_line(aes(color=sample)) +
  scale_x_continuous(limits = c(5,40)) +
  scale_y_continuous(limits = c(0,500)) +
  theme_bw()

# todo: add sliders for the y and x offsets
#cleanup
rm(df)
rm(files)
#
#
# play around with the manipulate package, 
# this works:)
manipulate(
      {df2 %>% filter(grepl(substrate, sample)) %>% 
                ggplot(aes(x = time, y = shift * as.numeric(as.factor(sample)) + sig)) +
                geom_line(aes(color=sample)) +
                scale_x_continuous(limits = c(xmin, xmax)) +
                scale_y_continuous(limits = c(0,ymax)) +
                theme_bw()}, #here ends the plotting func and come the manipulate arguments
    shift = slider(0,200, step=10),
    xmin = slider(0,20, step=1, initial = 0),
    xmax = slider(30,70, step = 2, initial = 70),
    ymax = slider(100, 1000, step = 10, initial = 1000),
    substrate = picker("PGA" = "PGA|mix", "Pectin" = "Pectin|mix", "all" = "*", initial = "all"))

#
#
readline(prompt = "Check out both the Plots and Viewer tabs, hit [enter] to end script. \nCheers!")

[1] ""

#end of script