editor_options
markdown wrap sentence

prolfquapp: Generating Dynamic DEA Reports using a command line interface to the prolfqua R Package

Read on bioRxiv https://www.biorxiv.org/content/10.1101/2024.10.09.617391v1 "prolfquapp - A User-Friendly Command-Line Tool Simplifying Differential Expression Analysis in Quantitative Proteomics"

Welcome to prolfquapp on GitHub! Here, you'll find everything you need to elevate your protein differential expression analysis. Prolfquapp integrates powerful preprocessing methods and advanced statistical models from the prolfqua R package prolfqua doi.org/10.1021/acs.jproteome.2c00441 to deliver insightful, clear visualizations and robust data outputs. Generate dynamic HTML reports, versatile file formats, and dive into interactive data visualization with ExploreDE. Prolfquapp implements a command line interface to run protein differential expression analysis, that can be integrated into your workflow manager.

Differential Expression Analysis Workflow with prolfquapp

After running your Quantification software, DIA-NN, MAXQUANT, FragPipe-TMT, FragPipe-DIA or FragPipe-LFQ, the quantification results are in an data_dir. Please add the .fasta file which was used by the quantification software to the data_dir.

prolfquapp is a set of command line tools.

To use it open your shell (linux, mac, or the linux subsystem on windows). Change into the directory with the identification/quantification results coming from FragPipe, MaxQuant, DIA-NN, Spectronaut etc. In the directory (in the terminal Linux or MAC) execute:

R --vanilla -e "prolfquapp::copy_shell_script(workdir = '.')" # if you have R on you system

or, if you have Docker on you system, you can also download prolfquapp_docker.sh and execute:

prolfquapp_docker.sh R --vanilla -e "prolfquapp::copy_shell_script(workdir = '.')" # if you wan to use the prolfqua docker image and container

This will place the following four shell script files (linux), or bat files (windows) into your working directory:

[1] "/<working_directory>/prolfqua_dea.sh"
[3] "/<working_directory>/prolfqua_yaml.sh"
[4] "/<working_directory>/prolfqua_qc.sh"
[5] "/<working_directory>/prolfqua_dataset.sh"

On Linux give the executables LINUX permissions:

chmod a+x prolfqua_*

All scripts can be run with the option --help.

Workflow Overview

1. Create Dataset
2. Generate Quality Control (QC)
3. Generate prolfqua YAML
4. Run Differential Expression Analysis

1. Create Dataset

The first step involves preparing the dataset by providing the experiment annotation. This is done using the prolfqua_dataset.sh script.

• Input: directory containing identification/quantification software outputs
• Output: csv, tsv or xlsx file template,

To create a prolfquapp compatible experiment annotation file run:

./prolfqua_dataset.sh -i data_dir/ -s DIANN -d annotation.xlsx # if R and prolfuqapp is installed on your computer

or if you use our prolfquapp docker container

./prolfquapp_docker.sh prolfqua_dataset.sh -i data_dir/ -s DIANN -d annotation.xlsx # if R you use the prolfquapp docker container

The annotation.xlsx file will be generated, and will contain 5 columns.

• Relative.Path/Path/raw.file/channel/ (unique*)

• name - used in tables and figures (unique*)

• group/experiment/ - main factor

• subject/bioreplicate (optional** or keep cells empty) - blocking factor

• control - used to specify the control condition (C) (optional)

• The rows must contain a unique value (no duplicates per column) ** If the experiment is not paired, or has no blocking factor (e.g. batch, cell line) delete the subject column.

The column raw.file is already filled out, based on the information available in the input directory. You will need to fill out the missing columns, e.g. group, subject, control. The column names are not case sensitive.

The annotation.xlsx file will be generated, and you will need to fill out the empty cells.

2. Generate Quality Control (QC)

The prolfqua_qc.sh script will create a QC report. The report consists of two HTML documents and XLSX file.

• Input: Dataset from step 1 and directory containing identification/quantification software outputs
• Output: QC report and visualizations

./prolfqua_qc.sh -i data_dir/ -p ProjectName -O ordername -w WorkunitName -d annotation.xlsx -s DIANN -o where_to_write_results
# or prefix with ./prolfquapp_docker.sh to run the docker container
# ./prolfquapp_docker.sh prolfqua_qc.sh -i data_dir/ -p ProjectName -O ordername -w WorkunitName -d annotation.xlsx -s DIANN -o where_to_write_results

This will generate a subfolder which starts with "QC_" with all the analysis results.

3. Generate prolfquapp yaml

Using the ./prolfqua_yaml.R command line tool you can set the parameters of the DEA and create a configuration file in YAML format.

• Output: Yaml file

./prolfqua_yaml.sh -y config.yaml
# or prefix with ./prolfquapp_docker.sh to run the docker container
# ./prolfquapp_docker.sh prolfqua_yaml.sh -y config.yaml

To see which parameters can be set using prolfqua_yaml.sh use the -h switch. Other parameters you can set by editing the yaml file.

4. Run Differential Expression Analysis

Finally, the prolfqua_dea.sh script runs the differential expression analysis using the configuration file generated in the previous step.

• Input: directory containing identification/quantification software outputs, dataset from step 1, configuration file from step 3.
• Output: folder containing differential expression analysis results (html files, excel tables, rank files, SummarizedExperiment.rds).

After setting the parameters in the config.yaml file you can run the DEA analysis by:

./prolfqua_dea.sh -i data_dir/ -d annotation.xlsx -y config.yaml -w NameOfAnalysis -s DIANN
# and again you run the version within the docker container with
# ./prolfquapp_docker.sh prolfqua_dea.sh -i data_dir/ -d annotation.xlsx -y config.yaml -w NameOfAnalysis -s DIANN

This will generate a subfolder which starts with "DEA_" and writes all the analysis results as well as the input data.

How to install

Linux

export R_LIBS_SITE="/scratch/PROLFQUA/r-site-library/"
R --vanilla << EOF
.libPaths()
install.packages(c("remotes","seqinr", "prozor", "logger"), repos = "https://stat.ethz.ch/CRAN/")
remotes::install_gitlab("wolski/prolfquadata", host="gitlab.bfabric.org")
remotes::install_github("fgcz/prolfqua", build_vignettes = TRUE, dependencies = TRUE)
remotes::install_github("prolfqua/prolfquapp", dependencies = TRUE)
EOF

Docker

ASMS poster: Streamlining Protein Differential Expression Analysis in Core Facilities

Related software

• Einprot https://github.com/fmicompbio/einprot
• LFQAnalyst https://analyst-suite.monash-proteomics.cloud.edu.au/apps/lfq-analyst/ and https://github.com/MonashBioinformaticsPlatform/LFQ-Analyst
• POMAShiny https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1009148
• MSDap https://github.com/ftwkoopmans/msdap