| Title: | Deconvolute Luminescence Readings on Bacterial Culture Plates |
|---|---|
| Description: | Implements the deconvolution algorithm developed in Mauri, Vecchione, and Fritz (2019) which enables deconvolution of luminscence readings for experimental culture plates. {reluxr} provides functions for calculating the 'best' deconvolution matrix from a calibration plate, and enables usage of this calibration matrix (or one calculated previously) to adjust experimental values from a plate reader. |
| Authors: | Brady Johnston [aut, cre] |
| Maintainer: | Brady Johnston <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.1.1 |
| Built: | 2025-03-14 02:59:06 UTC |
| Source: | https://github.com/bradyajohnston/reluxr |
Deconvolute a Single Vector
.decon_vec(vec, mat_decon).decon_vec(vec, mat_decon)
vec |
Numeric vector, representing the wells. |
mat_decon |
Deconvolution matrix created through |
a numeric vector, the same length as vec.
Deconvolute a Multi-Frame Matrix
.deconvolute_multi_frame_matrix(mat_frames, mat_decon).deconvolute_multi_frame_matrix(mat_frames, mat_decon)
mat_frames |
A multi-frame matrix, with each time point a row, and each column a well. |
mat_decon |
A deconvolution matrix created through |
A deconvoluted multi-frame matrix.
Expand and calculate a bleed-through matrix from the given matrix around the given reference well.
.mat_calc_bleed( mat, ref_row, ref_col, b_noise = 10, relative = TRUE, .f = mean ).mat_calc_bleed( mat, ref_row, ref_col, b_noise = 10, relative = TRUE, .f = mean )
mat |
A matrix to expand and calculate relative bleed-through based on a reference well. |
ref_row |
Reference well row number. |
ref_col |
Reference well column number. |
b_noise |
Calculated background noise to fill the bleed-through matrix. |
relative |
TRUE / FALSE whether to return relative values. |
.f |
Function to apply to the expanded matrix, defaults to mean. |
a matrix, that is a bleed-through matrix calculated from the given matrix.
Make Deconvolution Matrix From Bleed-Through Matrix
.mat_calc_deconvolution(mat).mat_calc_deconvolution(mat)
mat |
Bleed-through matrix from |
a matrix, ready for deconvolution.
Creats a matrix with each row being a time point in a multi-frame experiment, and each column is a single well. The columns are concatenated by row (ordered A1, A2, ... B1, B2, ...).
.multi_frame_matrix_from_df( data, value, time = "time", well = "well", arrange = FALSE ).multi_frame_matrix_from_df( data, value, time = "time", well = "well", arrange = FALSE )
data |
A dataframe with columns for the value, time and well ID. |
value |
Name of the column which contains the values for the matrix. |
time |
Name of the column which defines the time points for the frames (rows) of the matrix. |
well |
Name of the column which contains the well IDs of the samples. |
arrange |
Logical, whether to return a dataframe arranged by time and well. |
a matrix, with
Takes a multi-frame matrix and converts to a vector, with each frame being concatenated end-to-end. Each row / frame of the matrix is a time point, with each column representing a well (ordered A1, A2, ... B1, B2, ...).
.multi_frame_matrix_to_vec(mat, rowwise = TRUE).multi_frame_matrix_to_vec(mat, rowwise = TRUE)
mat |
A multi-frame matrix. |
rowwise |
Logical, whether to concatenate the rows (default, TRUE) or the columns. |
A vector, with values concatenated from the matrix.
Reorders a dataframe by the time points, then by the rows then by the columns.
df_arrange(data, time = "time", well = "well")df_arrange(data, time = "time", well = "well")
data |
A dataframe with a time column and a well column. |
time |
The name of the column with the time points. |
well |
The name of the column with the well ID information. |
The input dataframe reordered.
Logical test for well ID format.
is_well_id(x)is_well_id(x)
x |
A string vector. |
A logical vector.
is_well_id(c("a12", "a2", "a02", "foo1"))is_well_id(c("a12", "a2", "a02", "foo1"))
Read the output of Tecan Plate Readers
plate_read_tecan(file, temp = FALSE)plate_read_tecan(file, temp = FALSE)
file |
File path to the |
temp |
Logical, whether to include the temperature column. |
a tibble::tibble() of the values from the file.
fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) datfl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) dat
Using a deconvolution matrix, created through rl_calc_decon_matrix(),
adjust the values in the col_value column to take into account
bleed-through from surrounding wells.
rl_adjust_plate(data, value, mat_decon, time = "time", well = "well")rl_adjust_plate(data, value, mat_decon, time = "time", well = "well")
data |
A data frame that contains the experimental data. |
value |
The name of the column containing the values (i.e. 'lum'). |
mat_decon |
A deconvolution matrix created through
|
time |
The name of the column containing the time values (i.e. 'time') |
well |
Name of the column with the well ID information. |
A dataframe with the specified column having been deconvoluted, using the supplied deconvolution matrix.
fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) mat_d_best <- dat |> dplyr::filter(signal != "OD600") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix(value, time_s, ref_well = "E05", b_noise = 30) dat |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) ) dat |> dplyr::filter(signal == "LUMI") |> rl_adjust_plate(value, mat_d_best, time = time_s) |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) )fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) mat_d_best <- dat |> dplyr::filter(signal != "OD600") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix(value, time_s, ref_well = "E05", b_noise = 30) dat |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) ) dat |> dplyr::filter(signal == "LUMI") |> rl_adjust_plate(value, mat_d_best, time = time_s) |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) )
Use the data from a calibration plate, where the plate is empty except for a single well with a luminescent signal, to create a deconvolution matrix that can be used to adjust other experimental results.
rl_calc_decon_matrix( data, value, b_noise, time = "time", ref_well = "I05", well = "well" )rl_calc_decon_matrix( data, value, b_noise, time = "time", ref_well = "I05", well = "well" )
data |
A data frame that contains the data of the calibration plate. |
value |
Name of the column containing the luminescent values. |
b_noise |
The value of the background noise, which is the average signal for the background wells that are far away from the reference well. |
time |
Name of the column with the time values. |
ref_well |
The well ID of the reference well (i.e. 'E05', 'I12") |
well |
Name of the column with the well ID values. |
The deconvolution matrix will be unique for each plate type and plate-reader, so a matrix should be calculated for each combination of plate and plate reader, but once this is calculated, it can be re-used to adjust future experimental results.
a deconvolution matrix, for use in rl_adjust_plate()
fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) dat mat_d <- dat |> dplyr::filter(signal != "OD600") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix(value, time_s, ref_well = "E05", b_noise = 30) image(log10(mat_d))fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) dat mat_d <- dat |> dplyr::filter(signal != "OD600") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix(value, time_s, ref_well = "E05", b_noise = 30) image(log10(mat_d))
Calculate the Optimal Deconvolution Matrix
rl_mat_decon_best(mat, ref_row, ref_col, b_noise = 20)rl_mat_decon_best(mat, ref_row, ref_col, b_noise = 20)
mat |
A multi-frame matrix, where each row is a frame and each column is a well. |
ref_row |
Row number for the reference well. |
ref_col |
Column number for the reference well. |
b_noise |
Value for the background noise. |
a matrix, the optimised deconvolution matrix.
Plot a Plate-Layout of the Value Column
rl_plot_plate(data, value, well = "well", trans = log10)rl_plot_plate(data, value, well = "well", trans = log10)
data |
Dataframe with the value column and a column specifying the well ID. |
value |
Name of the column containing the value information to be displayed. |
well |
Name of the column with the well ID information for formatting the plate layout. |
trans |
Name of the transformation to apply to the data. Defaults to
|
a ggplot2::ggplot() plot.
fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) mat_d_best <- dat |> dplyr::filter(signal == "LUMI") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix("value", "time_s", ref_well = "E05", b_noise = 30) dat |> dplyr::filter(signal == "LUMI") |> dplyr::filter(time_s > 500) |> rl_adjust_plate(value, mat_d_best, time = time_s) |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) )fl <- system.file( "extdata", "calibrate_tecan", "calTecan1.xlsx", package = "reluxr" ) dat <- plate_read_tecan(fl) mat_d_best <- dat |> dplyr::filter(signal == "LUMI") |> dplyr::filter(time_s > 500) |> rl_calc_decon_matrix("value", "time_s", ref_well = "E05", b_noise = 30) dat |> dplyr::filter(signal == "LUMI") |> dplyr::filter(time_s > 500) |> rl_adjust_plate(value, mat_d_best, time = time_s) |> dplyr::summarise(value = mean(value), .by = well) |> rl_plot_plate(value, trans = log10) + ggplot2::scale_fill_viridis_c( limits = c(1, NA) )