A computational framework to study sub-cellular RNA localization

Area of Science:

• Cell Biology
• Molecular Biology
• Bioinformatics

Background:

• RNA localization is vital for cellular processes.
• Quantitative analysis of RNA distribution is challenging.
• Single molecule FISH (smFISH) enables RNA visualization.

Purpose of the Study:

• To develop an integrated analysis framework for sub-cellular RNA localization.
• To create robust features for quantifying diverse RNA localization patterns.
• To enable accurate classification of RNA localization using machine learning.

Main Methods:

• Development of an integrated analysis framework.
• Design and validation of RNA localization features using simulated images.
• Application of supervised and unsupervised learning for classification.
• Validation on experimental single molecule FISH (smFISH) data.

Main Results:

• A set of validated features describing RNA localization patterns (e.g., polarity, foci, membrane/nuclear association).
• Features are robust to RNA levels and applicable across cell lines.
• High accuracy classification of RNA localization patterns achieved.
• Demonstrated utility in measuring localization changes in perturbation experiments.

Conclusions:

• The developed framework provides accurate and robust quantitative analysis of RNA localization.
• RNA localization exhibits significant heterogeneity at the single-cell level.
• This suggests a dynamic and plastic nature of RNA localization in cells.