Real-time prediction of cell division timing in developing zebrafish embryo

Area of Science:

• Developmental biology
• In vivo imaging
• Computational biology

Background:

• Live-imaging and manipulation of developing embryos are crucial for studying developmental processes.
• Cell selection for in vivo manipulation typically relies on observer experience.
• A computational approach could enhance the accuracy and efficiency of cell selection.

Purpose of the Study:

• To develop a computer-assisted live-prediction method for identifying and selecting target cells for in vivo manipulation.
• To predict cell division timing in V2 neural progenitor cells in developing zebrafish embryos based on their shape changes.

Main Methods:

• Utilized 4D live-imaging data of developing zebrafish embryos.
• Developed a mathematical model to describe V2 cell geometry and its changes over time.
• Applied sequential Bayesian inference to predict cell division timing based on extracted shape features.

Main Results:

• Successfully predicted the division timing of individual V2 neural progenitor cells in real-time.
• Demonstrated the feasibility of using cellular shape dynamics for predictive modeling.
• Validated the method on randomly selected cells during live imaging.

Conclusions:

• The developed computer-assisted method can facilitate the identification and selection of target cells for in vivo live-manipulation.
• This predictive system offers a new opportunity for advancing in vivo experimental systems in developmental biology.
• Integrating computational prediction with live imaging enhances the precision of developmental studies.