Skip to content

This repository contains the code used in our benchmark of imputation tools for flow cytometry data.

Notifications You must be signed in to change notification settings

AUMC-HEMA/imputation-manuscript

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Introduction

This Github repository contains all code necessary for reproducing the analysis for our manuscript about imputation of MFC data. In general, all imputation and pre-processing was performed in R after which Jupyter Notebooks (Python) were used to visualize and interpret the resulting imputed FCS files. The following sections will explain which scripts were used for what purpose.

Pre-processing and imputation (R)

Transformation and generation of pseudo-tubes

Pre-processing and splitting of pre-gated (T-cell) data was performed using the script Split_MFC.R. For the real-life data, the Preprocess_MFC_RL.R was used. For Mosmann and Nilsonn datasets, the Split_Mosmann.R and Split_Nilsson.R files were used.

In this script, pre-gated FCS files are transformed with split into pseudo-tubes. This results in 3 different FCS files with either the suffix _gt (original, transformed data), _ff1 (pseudotube 1) or _ff2 (pseudotube 2).

Imputation of pseudo-tubes

Imputation of pseudo-tubes is performed in the script Impute_MFC.R. For the other datasets, please find the other R scripts which are named after the respective dataset.

Please account for the fact that this script does not impute for Infinicyt and requires these files to be already present beforehand.

The script CytoBackBone_MOD.R contains a modified version of the merge function which is used in Impute_MFC.R.

Generation of gating files

Gating was performed on aggregated files of ground truth and imputed data. These were generated using the script Create_MFC_Aggregates.R

Generation of gating labels

The script Get_MFC_Gating.R uses FlowJo workspace files in combination with the aggregated fcs files to generate csv files containing the gating labels.

Where to find the code for every plot in the manuscript

All analyses were performed in Jupyter notebooks.

Figure 2A, 2B, 2C: Distance_Plots.ipynb

Figure 3A, 3B, 3C: Density_Plots.ipynb

Figure 3D, 3E: Distance_Plots.ipynb

Figure 4: Backbone_Experiment.ipynb

Supplemental Figure 3: Distance_Plots.ipynb

Supplemental Figure 4: Infinicyt_Plots.ipynb

Supplemental Figure 5: Distance_Plots.ipynb

Supplemental Figure 6: Distance_Plots.ipynb

Supplemental Figure 7: Distance_Plots.ipynb

Supplemental Figure 8: Labeling_Plots.ipynb

Supplemental Figure 9: Labeling_Plots.ipynb

Supplemental Figure 10: FlowSOM_analysis.ipynb

Supplemental Figure 11: Backbone_experiment.ipynb

About

This repository contains the code used in our benchmark of imputation tools for flow cytometry data.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published