LysoSense CPS Analyzer

🌐 Hosted app: lysosense.streamlit.app · 📖 User guide: docs/USER_GUIDE.md

LysoSense is a reproducible workflow for analyzing differential centrifugal sedimentation (DCS/CPS) traces from E. coli homogenisation campaigns. It parses instrument .dat exports, fits bi-peak Gaussian/lognormal models to quantify intact cells and inclusion bodies, and serves interactive overlays, metrics, and downloadable summaries through a Streamlit app.

The data-processing strategy is adapted from the method described in Klausser et al., 2025.

Use the web app

• Open the hosted app: lysosense.streamlit.app
• Upload one or more CPS .dat files, inspect overlays and component fits, and download the XLSX summary.
• First time? Open the in-app 📖 Guide in the sidebar for a walkthrough with example plots (the written reference is docs/USER_GUIDE.md).

Features

• Parse CPS/DCS .dat exports into particle_size_um vs mass_signal_ug
• Constrained bi-peak fitting (intact cells vs inclusion bodies) with single-peak fallback, overlap deconvolution, and area-robustness tagging
• Metrics: component areas, intact fraction, lysis efficiency, mean sizes, R²
• Custom peak names, multiple peak models, and adjustable detection sensitivity
• Interactive Plotly overlays, a results table, and XLSX exports (summary + experimental data)
• A fully offline-capable in-app guide with example plots

Run locally (for development)

python -m venv .venv
.\.venv\Scripts\activate
python -m pip install -r requirements.txt
streamlit run app\streamlit_app.py

The app entry point adds src to sys.path at import time, so no editable install or PYTHONPATH setup is required.

Project layout

app/                  Streamlit entry point, in-app guide, and logo assets
src/lysosense/        Package: io.py (parsing), analysis.py (fitting)
tests/                pytest suite
docs/                 User guide
.github/workflows/    CI (ruff, mypy, pyright on Python 3.13)
requirements.txt      Pinned runtime dependencies

Citation

If this tool supports your work, please cite both the software and the underlying method:

• LysoSense (this repository) — use the "Cite this repository" button on the GitHub page (generated from CITATION.cff).
• Klausser et al., 2025. "Increased purity and refolding yield of bacterial inclusion bodies by recursive high pressure homogenization." Link.

Acknowledgements

Developed in the CD Laboratory for Inclusion Body Processing 4.0, IBD Group — Integrated Bioprocess Development, TU Wien. Funded by the Christian Doppler Gesellschaft and Boehringer Ingelheim.

License

Proprietary — © 2025–2026 Florian Gisperg and TU Wien. All rights reserved. See LICENSE. No part of this repository may be copied, modified, distributed, or used without prior written permission of the copyright holders.

---

Questions or feedback? Open an issue or submit a pull request.