FociPy is an open-source Python pipeline for the automated analysis of DNA repair protein recruitment kinetics at laser-induced damage sites. It processes timelapse fluorescence microscopy data from microirradiation experiments, extracts normalized intensity curves, and exports publication-ready quantitative results.
Designed for analyzing XRCC1 (single-strand break marker) and 53BP1 (double-strand break marker) accumulation in HeLa cells, but the modular architecture allows adaptation to other fluorescently-tagged repair proteins.
This software is being developed as part of a master's thesis in Molecular and Cellular Biophysics at the Jagiellonian University (Faculty of Biochemistry, Biophysics and Biotechnology).
Warning
UNDER DEVELOPMENT – NOT YET USABLE
FociPy is currently in the architecture design phase. The module structure is defined, but core algorithms are not yet implemented. Do not use for quantitative analysis.
In microirradiation experiments, a focused laser beam (488 nm) induces localized DNA damage within a single cell nucleus. Fluorescently-tagged repair proteins (e.g. GFP-XRCC1) are recruited to the damage site, forming a visible accumulation ("focus") that can be tracked over time. Traditional analysis in ImageJ/FIJI relies on manual ROI selection — a process that is subjective, time-consuming, and difficult to scale. FociPy will be a fully automated extraction of recruitment kinetics from raw microscopy files, using the normalization formula:
where:
-
$I_{foci}(t)$ — mean intensity within the repair focus ROI at time$t$ -
$I_{nuc}(t)$ — mean intensity of the entire nucleus (reference pool) -
$I_{bg}$ — mean background intensity (outside all nuclei)
This normalization corrects for variable expression levels (transient transfection) and photobleaching effects.
- Automated segmentation of cell nuclei from DAPI/Hoechst-stained timelapse images
- Tracking of laser-induced repair foci through timelapse sequences (center-of-mass drift correction)
- Extraction of normalized recruitment kinetics curves
- Quantitative analysis: half-recruitment time (
$t_{1/2}$ ), maximum intensity ($I_{max}$ ), plateau level - Investigation of dose-dependent DNA damage response (488 nm laser power vs. recruitment dynamics)
- Batch processing with full audit trail (source file hashes, pipeline parameters, QC images)
| Parameter | Value |
|---|---|
| Cell line | HeLa (transient transfection) |
| Markers | GFP-XRCC1 (SSB), GFP-53BP1 (DSB) |
| Microscope | Leica TCS SP5 FLIM |
| Objective | HCX PL APO 63x/1.40 OIL |
| Damage induction | Point microirradiation, 488 nm laser |
| Nuclear stain | DAPI / Hoechst |
| Data format | Leica .lif (native) |
| Module | Purpose |
|---|---|
config.py |
Pipeline parameters (dataclass + YAML) |
io_utils.py |
Image loading (.lif, .tif), metadata extraction, file hashing |
preprocessing.py |
Noise reduction, background subtraction |
segmentation.py |
Nuclear segmentation (Otsu+Watershed, Cellpose) |
kinetics.py |
ROI tracking, intensity measurement, normalization |
quantification.py |
Kinetic curve statistics (t_half, I_max) |
reporting.py |
CSV/JSON export, experiment metadata, audit trail |
visualization.py |
Kinetics plots, segmentation overlays, QC panels |
pipeline.py |
Orchestration of the full workflow |
| Category | Tools |
|---|---|
| Core | Python 3.10+, NumPy, Pandas, SciPy |
| Image processing | scikit-image, tifffile |
| Leica file support | readlif |
| Segmentation (optional) | Cellpose |
| Visualization | Matplotlib, Napari (QC) |
| Configuration | PyYAML, dataclasses |
- Project stage: architecture design
- Module structure: defined
- Core implementation: not started
- Validation: not started
This project is a research prototype developed for academic purposes (master's thesis). No guarantees are made regarding correctness, completeness, or suitability of the results at the current development stage.
Detailed documentation, including usage instructions (USAGE.md), methodology (METHODS.md), and changelogs (CHANGELOG.md), can be found in the docs/ directory.