pcnaDeep: a fast and robust single-cell tracking method using deep-learning mediated cell cycle profiling.

Area of Science:

• Cell Biology
• Bioimaging
• Computational Biology

Background:

• Single-cell analysis is crucial for understanding cellular decisions.
• Robust tracking of cells through division is a major challenge in biomedical research.
• Existing methods lack efficiency in long-term single-cell lineage tracing.

Purpose of the Study:

• To develop an automated application for high-fidelity, long-term single-cell tracking.
• To enable accurate assignment of mother-daughter cell relationships.
• To facilitate advanced cell cycle analysis in time-lapse microscopy.

Main Methods:

• Integration of a deep-learning-based instance segmentation module for Proliferating Cell Nuclear Antigen (PCNA) signal detection.
• Development of a cell tracking and cell cycle resolving pipeline.
• Incorporation of fluorescent cell cycle reporter information to identify mitosis events.

Main Results:

• An application that automatically tracks single cells and assigns mother-daughter relationships.
• High-fidelity long-term single-cell tracking achieved by associating mitosis events.
• User-friendly interface and extensible APIs for customized analysis and manual correction.

Conclusions:

• The developed application significantly improves the ability to perform long-term single-cell tracking.
• This tool addresses a key barrier in single-cell analysis for biomedical research.
• pcnaDeep offers a robust solution for studying cell division and lineage in microscopy studies.