In MartinBanchero/mpxtractor: Process microplate reader output files into a tidy dataframe

library(mpxtractor)
library(dplyr)
options(width = "95")
knitr::opts_chunk$set(collapse = TRUE, comment = "#>", fig.align = "center", fig.show = "hold", dev = "png")

Advantages of using mpxtractor

It is well known for anyone performing experiments that involved microplate readers that the analysis of the raw data can be tedious and time-consuming. Several of this microplate reader machines include software to alleviate this problem but many times shows a lack of standardization. Although several good R packages exist many of them are design to solve specific problems losing generability. For example, some features from @plater were used to build mpxtractor. To clarify, mpxtractor can process raw data from three different microplate reader machines, SpectraMax, MultiscanGO and FluorStar.

For the reasons mentioned above is why mpxtractor can help to reduce the time to process the raw data into tidy data frames, combine layout files with tidy data. Also, mpxtractor provides features to visualize growth curves and layout files over a microplate shape. The latter can be printed to be used in the lab during the setups of experiments.

Features of mpxtractor

This package is divided into two sets of functions

1. Function to wrangling the data in raw files, these functions are explained in detail in vignette wrangling_and_layout_functions alongside with some examples.

2. Functions to plot layout files and growth curves, the vignette plotting_functions explain how they work, and some details.

Example

Layout design and visualization

Let's suppose we want to experiment to measure growth rates under different types of sugars. The first step is to design the layout and generate the layout file. Using mpxtractor we can plot the layout file. or load it into a tidy data frame. Now we want to set the microplates and run the experiment in the lab, for this we can use the plot_layout_file() function to have a visualization of our experimental design. For example, to avoid edge effects due to temperature, evaporation, etc. the wells around the perimeter of the plate might be more sensitive to these effects, visualizing the layout design over the microplate is easier to check whether the control or the samples are well positioned in the plate. In Fig.1 is shown the basic containing control or sample experiment while condition contains the type of sugar to be used in the experiment.

layout_file <- system.file("extdata", "test_layout_file.csv", package = "mpxtractor")

mpxtractor::plot_layout_file(file = layout_file, var_shape = "basic", var_colour = "condition", add_conc = TRUE, output_filename = "spectramax_layout.png" )

knitr::include_graphics('spectramax_layout.png')

As you can see in Fig.1 only two conditions can be add to this visualization, one given the shape (in this case basic) and one the color(condition). Is possible to load the layout file into a tidy data frame using read_layout_file().

layout_file <- system.file(
  "extdata",
  "test_layout_file.csv",
  package = "mpxtractor"
)

df_layout <- mpxtractor::read_layout_file(file = layout_file)

head(df_layout)

As we observed above we have several variables, basic,condition, strainand concentration. For example we can join conditions and obtain a more broad visualization.

# We have all the layout information in a tidy dataframe obtained from before.
df_layout <- mpxtractor::read_layout_file(file = layout_file)

df_unite <- tidyr::unite(
  df_layout,
  basic_strain,
  -c(Wells, condition, concentration)
)

head(df_unite)

Data wrangling

Once all the experimental design is complete we are ready to use the microplate reader machines. In this example, we are using a multiscanGO to measure our growth rates. Within mpxtractor we can find one wrangling function for each machine, in this case we going to use read_multiscango_data() (see vignette wrangling_and_layout_functions) to store the raw data into a tidy data frame.

Note: This function takes as argument time point to build the time series. In the example below the time_point (reading) is 2 minutes.

file <- system.file("extdata", "test_gr_spectramax.txt", package = "mpxtractor")

# Extract the data stored in the files into a data frame using proper wrangling function
df_spmax <- mpxtractor::read_spectramax_data(file = file)

# Show tidy data 
head(df_spmax) 

Now let's suppose we want to repeat the experiment under the same conditions. In this case, we have to raw data files to be analyzed. Using the function read_multiple_data_files() we can store both files into a tidy data frame. There many options that can be used to gather files with this function, for more information check vignette wrangling_and_layout_functions .

# Path to files
file1 <- system.file("extdata",
  "test_gr_spectramax.txt",
  package = "mpxtractor"
)

file2 <- system.file("extdata",
  "test_spectramax_data_2.txt",
  package = "mpxtractor"
)

df_two_files <- mpxtractor::read_multiple_data_files(
  reader_type = "spectramax",
  filesname = c(file1, file2),
  plate_names = c("plate_1", "plate_2")
)

head(df_two_files)
tail(df_two_files)

At this point, we have two measurements under the same layout design. Now we want to add the layout information to our tidy data frame with both files. To do this we can use the function combine_data_with_layout() and store all the information in one tidy data frame.

#get the path to layout file
file_layout <- system.file(
  "extdata",
  "test_layout_file.csv",
  package = "mpxtractor"
)

df <- mpxtractor::combine_data_with_layout(
  df_data = df_two_files, 
  reader_type = "spectramax", 
  layout_files = file_layout
  )
df

plt <- mpxtractor::plot_gr_microplate(
  df_data = df, 
  cond_to_col = "condition", 
  var_gr = "Measurement",  
  ws = "2 hs", 
  plate_file = "plate_1", output_filename = "growth_rates.png")

knitr::include_graphics('growth_rates.png')

References

MartinBanchero/mpxtractor documentation built on March 30, 2022, 10:56 p.m.