Skip to contents

TockyPrep: Data Preprocessing Methods for Flow Cytometric Fluorescent Timer Analysis

Dr. Masahiro Ono

2024-11-20

Source: vignettes/TockyDataPreprocessing.Rmd
TockyDataPreprocessing.Rmd

Introduction

TockyPrep: Standardizing the Analysis of Fluorescent Timer Proteins in Flow Cytometry

Fluorescent Timer proteins uniquely change their emission spectra over time and serve as powerful tools for monitoring the dynamic processes within cells. These proteins are crucial for understanding the intricate temporal dynamics of cellular events. Despite their potential, the analysis of Timer fluorescence data in flow cytometry is often complicated by variability in instrument settings and the absence of standardized data preprocessing methods. To overcome these challenges, the TockyPrep package has been developed. This R package provides a comprehensive suite of tools designed to automate the preprocessing, normalization, and trigonometric transformation of Timer fluorescence data.

Aim

The TockyPrep package aims to standardize the analysis of Timer fluorescence to improve reproducibility and accuracy across various experimental setups. It specifically addresses the normalization of immature and mature Timer fluorescence as a critical step for robust analysis. This approach is central to understanding the maturation dynamics of Timer proteins, and is implemented using advanced trigonometric transformations.

Getting Started with TockyPrep

To begin using TockyPrep, install the package from GitHub using the following command:

# Install TockyPrep from GitHub
devtools::install_github("MonoTockyLab/TockyPrep")

Once installed, you can load the package and follow the vignette’s sample workflow to process your flow cytometry data. This vignette will guide you through importing your data, applying preprocessing steps, and visualizing the results to ensure that you can fully leverage the capabilities of the TockyPrep toolkit.

Sample Workflow

This section will walk you through a typical analysis workflow using TockyPrep to process flow cytometric data of cells expressing Fluorescent Timer proteins. We will cover data import, application of preprocessing methods, and basic visualization techniques.

Set-up and prep_tocky

To start the TockyPrep analysis, load the library.

We will use example files included in the TockyPrep package. To access the files, run the following codes, which defines the sample files and the negative control file to be used.

# Example data load
# Define the base path
file_path <- system.file("extdata", package = "TockyPrep")

# Define files
negfile <- "Timer_negative.csv"
samplefiles <- list.files(file_path, pattern = "sample_", full.names = FALSE)
samplefiles <- setdiff(samplefiles, file.path(file_path, negfile))


print(negfile)
## [1] "Timer_negative.csv"
head(samplefiles)
## [1] "sample_0hrs_R1.csv"  "sample_0hrs_R2.csv"  "sample_0hrs_R3.csv" 
## [4] "sample_12hrs_R1.csv" "sample_12hrs_R2.csv" "sample_12hrs_R3.csv"

We will define the files to be analyzed using the function prep_tocky.

# Preprocessing data
prep <- prep_tocky(path = file_path, samplefile = samplefiles, negfile = negfile, interactive = FALSE)

Data preprocessing by timer_transform

Now we are ready to apply the data preprocessing methods to the sample data. timer_transform does all the steps, including the data import, Timer thresholding, Timer fluorescence normalization, and Trigonometric transformation to obtain Timer Angle and Intensity values for individual cells.

# Normalizing and transforming data
transformed_data <- timer_transform(prep, blue_channel = 'Timer.Blue', red_channel = 'Timer.Red', select = FALSE, verbose = FALSE)

Visualization

Preparation: sample grouping definition

To effectively visualize the processed data, sample grouping needs to be defined using sample_definition. The example file sampledefinition.csv provides the standard form. Use data.frame as input into sample_definition.

sample_definition <- read.csv(file.path(file_path, 'sampledefinition.csv'))
sample_definition <- as.data.frame(sample_definition)
head(sample_definition)
##                  file group
## 1  sample_0hrs_R1.csv     0
## 2  sample_0hrs_R2.csv     0
## 3  sample_0hrs_R3.csv     0
## 4 sample_12hrs_R1.csv    12
## 5 sample_12hrs_R2.csv    12
## 6 sample_12hrs_R3.csv    12
# Normalizing and transforming data

transformed_data <- sample_definition(transformed_data, sample_definition = sample_definition, interactive = FALSE)

Alternatively, sample grouping can be created by choosing interaction session using the option interactive = TRUE. This will prompt generation of sample_definition.csv in your output directory.

Confirming the gating of Timer fluorescence

To visualize the thereshold values for Timer blue and red fluorescence, use plot_timer_gating.

plot_timer_gating(prep = prep, x = transformed_data)
## [1] "FSC.A"      "Timer.Blue" "Timer.Red"
Figure 1: Thereshold values set for Timer blue and red fluorescence.

Figure 1: Thereshold values set for Timer blue and red fluorescence. 

## Percentage of cells in each quadrant:
## Q1: 0.2029%
## Q2: 0%
## Q3: 0.2029%
## Q4: 99.5941%

Visualization of the processed Timer data

To visualize the processed Timer data, use plot_tocky. This is a versatile function, enabling the three plot modes, including 'Timer fluorescence', 'Normalized Timer fluorescence', and 'Timer Angle and Intensity'.

  • 'Timer fluorescence': Timer blue and red fluorescence, without normalization.
  • 'Normalized Timer fluorescence': Normalized and thrsholded Timer blue and red fluorescence.
  • 'Timer Angle and Intensity': Timer Angle and Intensity after trigonometric transformation.

Non-normalized Timer fluorescence (Figure 2)

The default is 'Timer fluorescence', which is non-normalized Timer fluorescence.

logic <- sample_definition$group %in% c(4, 8, 12, 16)
samplefiles <- sample_definition$file[logic]
show(samplefiles)
##  [1] "sample_12hrs_R1.csv" "sample_12hrs_R2.csv" "sample_12hrs_R3.csv"
##  [4] "sample_16hrs_R1.csv" "sample_16hrs_R2.csv" "sample_16hrs_R3.csv"
##  [7] "sample_4hrs_R1.csv"  "sample_4hrs_R2.csv"  "sample_4hrs_R3.csv" 
## [10] "sample_8hrs_R1.csv"  "sample_8hrs_R2.csv"  "sample_8hrs_R3.csv"
# Visualizing the results
plot_tocky(transformed_data, interactive = FALSE, save = FALSE, n = 2, samplefile = samplefiles, verbose = FALSE)
Figure 2: Non-normalized Timer Fluorescence in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Figure 2: Non-normalized Timer Fluorescence in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Normalized Timer fluorescence (Figure 3)

Using the option plot_mode = "Normalized Timer fluorescence", you can visualize Normalized Timer blue and red fluorescence.

# Visualizing the results
plot_tocky(transformed_data, plot_mode = "Normalized Timer fluorescence", interactive = FALSE, save = FALSE, n = 2, samplefile = samplefiles, verbose = FALSE)
Figure 3: Normalized Timer Fluorescence in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Figure 3: Normalized Timer Fluorescence in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Timer Angle and Intensity (Figure 3)

Using the option plot_mode = "Timer Angle and Intensity", you can visualize the effects of the trigonometric transformation applied to the normalized data.

# Visualizing the results
plot_tocky(transformed_data, plot_mode = "Timer Angle and Intensity", interactive = FALSE, save = FALSE, n = 2, samplefile = samplefiles, verbose = FALSE)
Figure 3: Timer Angle and Timer Intensity in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Figure 3: Timer Angle and Timer Intensity in T cells stimulated by antigen and analyzed at 4, 8, 12, and 16 hours after activation.

Conclusion

TockyPrep provides a comprehensive suite of data preprocessing methods for deriving Timer Angle and Timer Intensity data. This suite includes Timer thresholding, Timer normalization, and trigonometric transformation. The resulting Timer Angle and Intensity data can be robustly used in downstream analyses.

Please explore downstream analysis examples in Bending, Martı́n, et al. (2018); Bending, Paduraru, et al. (2018); and Hassan et al. (2022).

References

Bending, David, Paz Prieto Martı́n, Alina Paduraru, Catherine Ducker, Erik Marzaganov, Marie Laviron, Satsuki Kitano, Hitoshi Miyachi, Tessa Crompton, and Masahiro Ono. 2018. “A Timer for Analyzing Temporally Dynamic Changes in Transcription During Differentiation in Vivo.” Journal of Cell Biology 217 (8): 2931–50. https://rupress.org/jcb/article/217/8/2931/39442/A-timer-for-analyzing-temporally-dynamic-changes.
Bending, David, Alina Paduraru, Catherine B Ducker, Paz Prieto Martín, Tessa Crompton, and Masahiro Ono. 2018. “A Temporally Dynamic Foxp3 Autoregulatory Transcriptional Circuit Controls the Effector Treg Programme.” Journal Article. The EMBO Journal 37 (16): e99013. https://www.embopress.org/doi/full/10.15252/embj.201899013.
Hassan, Jehanne, Elizabeth Appleton, Bahire Kalfaoglu, Malin Pedersen, José Almeida-Santos, Hisashi Kanemaru, Nobuko Irie, et al. 2022. “Single-Cell Level Temporal Profiling of Tumour-Reactive t Cells Under Immune Checkpoint Blockade.” bioRxiv. https://doi.org/10.1101/2022.07.19.500582.