Post-processing

Here we collect Jupyter notebooks, Quarto documents, and Marimo notebooks that investigate the ProteoBench data.

Examples include testing new and additional benchmarking metrics, parallel processing of uploaded data files, and manuscript-ready visualizations.

Contents

Submission Utilities

Tools for preparing and converting data before uploading to ProteoBench.

Notebook	Description
submission/ProteoBench_input_conversion.ipynb	Prepares ProteoBench inputs from software tool outputs where the required information is spread across more than one file.
utils/create_mapper.ipynb	Creates a protein description–to–gene name mapper CSV from FASTA files, used by the ProteoBench parsing layer.

Single-Submission Analysis

Notebooks for inspecting and visualizing results from an individual ProteoBench submission.

Notebook	Description
analysis/single_submission/quant/lfq/ion/dda/logFCs_per_organism_plots.ipynb	Demonstrates how to create figures from ProteoBench intermediate files. Reproduces an intensity-vs-logFC plot (similar to Fig. 3d in the ion-quantification benchmark publication by Van Puyvelde et al., 2022).

Post-Analysis — DDA Ion Quantification

Cross-submission analyses for data-dependent acquisition (DDA) ion-level label-free quantification modules.

Notebook / Document	Format	Description
analysis/post_analysis/quant/lfq/ion/dda/DDA_ion_quant_manuscript.ipynb	Jupyter	Generates manuscript figures for the DDA ion quantification module, including upset plots and intensity boxplot comparisons across software tools.
analysis/post_analysis/quant/lfq/ion/dda/indepth_module_analysis.ipynb	Jupyter	In-depth performance benchmarking: downloads all public DDA submissions, re-runs the ProteoBench scoring pipeline, and measures load, conversion, and benchmarking times per tool.
analysis/post_analysis/quant/lfq/ion/dda/indepth.qmd	Quarto	Quarto version of the in-depth module analysis above, suitable for rendering as a self-contained HTML report.
analysis/post_analysis/quant/lfq/ion/dda/scores_dissected.py	Marimo	Interactive accuracy-vs-precision analysis for DDA quantification: loads benchmark results from GitHub and produces per-species scatter and box plots.
analysis/post_analysis/quant/lfq/ion/dda/plot_runs_sage.ipynb	Jupyter	Downloads Sage search results from Figshare and plots metric performance across multiple Sage parameter configurations.

Post-Analysis — DIA Ion Quantification

Cross-submission analyses for data-independent acquisition (DIA) ion-level label-free quantification modules.

Notebook	Format	Description
analysis/post_analysis/quant/lfq/ion/dia/astral_diaPasef/Astral_diaPASEF_manuscript.ipynb	Jupyter	Manuscript figures for the Astral and diaPASEF DIA modules: downloads submissions from GitHub, re-scores them, and generates upset plots and metric scatter plots.

Multi-Module Overview

Notebook	Format	Description
analysis/post_analysis/quant/lfq/ion/benchmark_analysis.py	Marimo	Interactive benchmark dashboard supporting multiple DDA and DIA modules. Loads results from GitHub and renders per-species accuracy-vs-precision scatter plots with a configurable module selector.
analysis/post_analysis/quant/lfq/ion/index.py	Marimo	Index page listing all available ProteoBench benchmark analysis reports with submission counts per module and links to the individual HTML reports.

Exploratory Analysis

Notebook	Description
analysis/methionine_excision_analysis.ipynb	Analyses all publicly submitted ProteoBench datasets to determine which software tools apply N-terminal methionine excision (NME) and whether there is a systematic difference between tools (related issue #504).

Contributing

Contributions are very welcome! You can contribute:

• Jupyter notebooks (.ipynb) — the classic format for interactive analysis.
• Marimo notebooks (.py) — reactive Python notebooks that are version-control-friendly and can be run as scripts or apps.
• Quarto documents (.qmd) — literate programming documents that render to HTML, PDF, or other formats.

If you have an analysis, visualization, or utility script that builds on ProteoBench data, feel free to open a pull request. Please place your contribution in the most appropriate subdirectory (e.g. analysis/, submission/, utils/) and add a short description to this README.