Structural Repetition Detector for multi-scale quantitative mapping of molecular complexes through microscopy.

Area of Science:

• Cell biology
• Biophysics
• Computational biology

Background:

• Cells display recurring structural motifs across scales, offering functional insights.
• Manual detection of these structures in large microscopy datasets is difficult and prone to bias.

Purpose of the Study:

• To introduce the Structural Repetition Detector (SReD), an unsupervised computational framework for identifying repetitive biological structures.
• To provide an unbiased method for analyzing cellular patterns irrespective of imaging modality.

Main Methods:

• SReD exploits local texture repetition to formulate structure detection as a similarity-matching problem between image regions.
• The framework operates without prior knowledge of the structures or imaging techniques.

Main Results:

• SReD successfully identified repetitive patterns in various fluorescence microscopy images.
• Quantitative analyses demonstrated utility in estimating spectrin ring periodicity in neurons, detecting Human Immunodeficiency Virus type-1 assembly, and evaluating microtubule dynamics.

Conclusions:

• SReD offers an unsupervised and unbiased approach to detect and analyze repetitive biological structures.
• The open-source ImageJ/FIJI plugin facilitates diverse applications in cell biology research across different imaging modalities.