Skip to contents

Process experimental data (normalise, correct and calibrate)

Source: R/process_plate.R
process_plate.Rd

Processes experimental (fluorescence and absorbance) data from microplate readers. Expects 'parsed' data that is tidy and attached to appropriate metadata. Numerous arguments offer options to customise how and whether normalisation, correction and/or calibration is carried out. Data can be timecourse (kinetic) data or single timepoint (endpoint) data, and can contain fluorescence data only or include absorbance (OD) data to enumerate cell density. Originally based on flopr::process_plate() but with numerous changes. First, the function normalises optical density readings (if present) to media/buffer wells, and fluorescence readings, either to media/buffer wells or to cellular autofluorescence. Fluorescence channels must be specified in flu_channels, the blank wells in blank_wells, and the reading to be used for optical density in od_name. The cellular autofluorescence model can be set with af_model on the negative (neg_well) wells, or set to NULL. Second, if do_quench_correction is TRUE, the 'quenching' or attenuation of fluorescence measurements by cells is corrected for according to the cell density measured. This requires the specification of od_type as "OD600" or "OD700". Third, if do_calibrate is TRUE, normalised values for both fluorescence and optical density are converted into meaningful units. Here, the paths to conversion factor CSV files must be provided as od_coeffs_csv for OD and fluor_coeffs_csv for fluorescence. For fluorescence measurements, each calibration must be specified by flu_slugs to represent the FP used, flu_gains to provide the gain, and flu_labels to specify how the relevant plots should be labelled, (e.g. flu_slugs = c("mcherry", "mtagbfp2"), flu_gains = c(60,80), flu_labels = c("RFP, BFP")).

Usage

process_plate(
  data_csv,
  blank_well = "A1",
  timecourse = TRUE,
  od_name = "OD",
  flu_channels = c("green1green2"),
  flu_channels_rename = NULL,
  af_model = "spline",
  neg_well = "A2",
  do_quench_correction = FALSE,
  od_type,
  do_calibrate = FALSE,
  instr,
  flu_slugs = c(),
  flu_gains = c(),
  flu_labels = c(),
  od_coeffs_csv,
  fluor_coeffs_csv,
  outfolder = ".",
  csv_only = FALSE
)

Arguments

data_csv

path to a CSV file containing parsed plate reader data

blank_well

the well coordinates of one or more media blanks. Defaults to "A1".

timecourse

logical. Is the data timecourse/kinetic data and does it include a variable called 'time'?

od_name

the column name for the optical density data. Defaults to "OD". If no OD measurements were taken, use NULL.

flu_channels

the column names for the fluorescence data. Defaults to "green1green2".

flu_channels_rename

if specified, what to change the flu_channels column names to, before computing normalisations and calibrations. Can be useful if the channel_name in the calibration file is named differently from the columns in the data file. Needs to be same length as flu_channels, if not all require changing, specify them anyway to allow positional replacement (first element in flu_channels_rename replaces first in flu_channels, etc.). Defaults to NULL.

af_model

model used to fit negative control autofluorescence. For now these include "polynomial", "inverse_poly", "exponential", "spline" and "loess". If set to NULL, no model is used, and fluorescence is normalised akin to OD: by subtracting the value for the blanks. Defaults to "spline".

neg_well

the well coordinates of a non-fluorescent control. Defaults to "A2".

do_quench_correction

logical. Should function correct for anticipated quenching of fluorescence, depending on the cell density?

od_type

Which OD-type was used? Required for quench correction. "OD600" or "OD700".

do_calibrate

logical. Should function convert OD and fluorescence data to calibrated units? Defaults to FALSE.

instr

character string to represent instrument. If do_calibrate = TRUE, used for filtering od_coeffs_csv and fluor_coeffs_csv files for conversion factors of the relevant instrument.

flu_slugs

character array representing fluorescent proteins (format = FPbase slug). If do_calibrate = TRUE, used for filtering fluor_coeffs_csv for conversion factors of the relevant FP.

flu_gains

numeric array representing gains of each fluorescent channel. If do_calibrate = TRUE, used for filtering fluor_coeffs_csv for conversion factors of the relevant gain.

flu_labels

If do_calibrate = TRUE, the column names to be given to each calibration. May be identical to flu_slug or flu_channel, but recommended is to make it obvious which FP is being calibrated, e.g. "mCherry", as channel names may be non-specifically named e.g. "red1red1". Needs to be same length as flu_slugs and flu_gains.

od_coeffs_csv

if do_calibrate = TRUE, path of the CSV file containing conversion factors for optical density

fluor_coeffs_csv

if do_calibrate = TRUE, path of the CSV file containing predicted conversion factors for the fluorescent channels

outfolder

path to folder where output files should be saved. Defaults to current working directory.

csv_only

logical. Saves only CSV files as outputs when TRUE. Defaults to FALSE.

Value

a data.frame with columns for raw plate reader data, normalised data and, if do_calibrate = TRUE, calibrated OD and FP data

Examples

if (FALSE) {
  processed_data <- process_plate(
    data_csv = "mcherry_parsed.csv",
    blank_well = c("A11"),
    od_name = "OD600", flu_channels = c("red1"), flu_channels_rename = c("red1red1"),
    af_model = NULL,
    do_quench_correction = TRUE, od_type = "OD700",
    do_calibrate = TRUE, instr = "spark1",
    flu_slugs = c("mcherry"), flu_gains = c(80), flu_labels = c("mcherry"),
    od_coeffs_csv = "od_coeffs.csv", fluor_coeffs_csv = "flu_coeffs.csv",
    outfolder = file.path("data_processed")
  )
}