ProteoFlux

ProteoFlux is an open-source Python framework for transparent, reproducible downstream analysis of quantitative proteomics data.

It operates on quantitative outputs generated by upstream search and quantification tools and provides a unified workflow for:

• Protein-centric proteomics
• Peptide-centric analyses
• Phosphoproteomics (with optional protein-level covariate adjustment)
• PELSA analyses

ProteoFlux emphasizes:

• Explicit data semantics
• Transparent preprocessing
• Deterministic statistical modeling
• Fully reproducible outputs

---

Overview

ProteoFlux takes a single YAML configuration file and produces:

• Harmonized and preprocessed quantification data
• Differential expression results using a limma-based empirical Bayes framework
• Principal component analysis (PCA), multidimensional scaling (MDS), and hierarchical clustering
• A structured multi-page PDF report
• Portable .h5ad files compatible with ProteoViewer
• Summary tables in Excel

---

Installation

Python ≥ 3.9 required.

Install from source (recommended)

git clone https://github.com/afanc/proteoflux.git
cd proteoflux
pip install -e .

This installs ProteoFlux in editable mode.

---

Quickstart (CLI)

List available configuration templates:

proteoflux templates

Create a template:

proteoflux init spectronaut-proteomics --path config.yaml

Edit the file and then run the full pipeline:

proteoflux run --config config.yaml

---

Output Files

File	Description
*.h5ad	Full AnnData object with raw, normalized, imputed data, statistics, embeddings, and metadata
*.xlsx	Differential expression summary table
*.pdf	Multi-page QC and analysis report

The .h5ad files are compatible with ProteoViewer, an interactive visualization tool for ProteoFlux outputs.

---

Using ProteoFlux as a Python Library

ProteoFlux can also be used programmatically:

import yaml
from proteoflux.main import run_pipeline

config = yaml.safe_load(open("config.yaml")) #or use a dict
run_pipeline(config)

This allows:

• Integration into larger workflows
• Batch processing
• Use inside Jupyter notebooks
• Custom downstream analysis of AnnData objects

All behavior is controlled by the same YAML schema used by the CLI.

---

Configuration

All parameters are defined in a single YAML configuration file. We recommend starting with a template:

proteoflux templates
proteoflux init TEMPLATE_NAME

Full parameter reference:

See the configuration guide docs/CONFIGURATION.md.

---

Examples

Runnable, reduced example datasets and matching configs are provided under examples/:

• examples/searle_small/: small DIA proteomics example (Spectronaut export subset)
• examples/phospho_small/: phosphoproteomics example with an injected flow-through covariate run (Spectronaut export subsets)
• examples/pelsa_small/: small pelsa analysis example (Spectronaut export subsets)

Each example folder contains a README.md with a minimal command to run the pipeline (typically proteoflux run --config <config>.yaml) and the required input files (data + annotation, if applicable).

---

Paper

The manuscript source is provided as paper.md.

---

License

ProteoViewer is released under the MIT License. See the LICENSE file for details.

---

Citation

If you use ProteoFlux in your work, please cite:

ProteoFlux (latest version).
Zenodo. https://doi.org/10.5281/zenodo.18640998