MatCol

MatCol performs object-based colocalization analysis of fluorescence microscopy images to quantify and statistically assess co-localization of signals and address limitations of pixel intensity-based coefficients such as Pearson, Manders, Li, and Costes.

Key Features:

• Object-based colocalization: Identifies regions of fluorescent signal across two channels and determines co-located segments for object-level colocalization measurements.
• Statistical significance testing: Calculates statistical significance of observed colocalizations using Student's t-test by comparing observed overlaps to overlaps expected by random chance.
• Noise-reduction filters: Implements median and Wiener filtering to enhance signal-to-noise ratio prior to object detection.
• Batch processing: Supports processing of multiple images for high-throughput colocalization analysis.
• Validation and accuracy: Validated on telomeric DNA with TRF2 and TRF2 with PML proteins across more than 350 nuclei from three different cell lines, showing a linear regression correlation between manual and MatCol-identified colocalizations (R² = 0.81, P < 0.0001).

Scientific Applications:

• Protein–DNA interaction analysis: Quantifies co-localization between proteins (e.g., TRF2, PML) and DNA (e.g., telomeric DNA) to study protein–DNA associations.
• Subcellular localization studies: Measures spatial relationships of molecular complexes and cellular structures at the object level within fluorescence microscopy data.
• Quantitative validation of microscopy colocalization: Provides statistically validated object-level colocalization metrics for comparison with manual annotation and experimental controls.

Methodology:

Object-based detection of fluorescent regions across two channels; median and Wiener filtering for noise reduction; statistical significance assessment via Student's t-test comparing observed overlaps to random overlaps; linear regression used for validation (reported R² = 0.81, P < 0.0001).