A computational pipeline for image-based statistical analysis of biomolecular condensates dynamics using morphological descriptors
Abstract Biomolecular condensation has been extensively studied recently, yet advanced analytical methods for characterizing phase-separated systems remain limited. We developed a Python-based computational pipeline compatible with desktops and HPC systems that quantifies morphological heterogeneity...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-09148-y |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Biomolecular condensation has been extensively studied recently, yet advanced analytical methods for characterizing phase-separated systems remain limited. We developed a Python-based computational pipeline compatible with desktops and HPC systems that quantifies morphological heterogeneity in biomolecular condensates through a user-friendly Jupyter notebook platform. Our approach employs advanced morphological features, including Euler characteristic number and fractal dimension, to describe subtle spatiotemporal information from biomolecular condensates. We implemented robust statistical analyses besides conventional descriptors, incorporating skewness and kurtosis for asymmetric data distribution, and multivariate analysis through interactive principal component analysis (PCA) visualization combined with correlation and scree plots. The proposed statistical framework was applied to study the condensation of the neurodevelopmental protein DDX3X, which assembles spherical droplets in-vitro. The presence of PEG3350 as a crowding agent induced the formation of condensates with altered morphology. Furthermore, the R376C mutant, linked to neurodevelopmental disorders, formed elongated aggregates over time. Our pipeline’s tools for live plotting, phase diagram analysis, and high-throughput automation enable detailed investigation of condensate assembly dynamics, advancing standardization of morphological descriptor analysis of biomolecular condensates. |
|---|---|
| ISSN: | 2045-2322 |