Challenges in Imaging Analyses of Biomolecular Condensates in Cells Infected with Influenza A Virus.

Biomolecular condensates are crucial compartments within cells, relying on their material properties for function. They form and persist through weak, transient interactions, often undetectable by classical biochemical approaches. Hence, microscopy-based techniques have been the most reliable methods to detail the molecular mechanisms controlling their formation, material properties, and alterations, including dissolution or phase transitions due to cellular manipulation and disease, and to search for novel therapeutic strategies targeting biomolecular condensates. However, technical challenges in microscopy-based analysis persist. This paper discusses imaging, data acquisition, and analytical methodologies' advantages, challenges, and limitations in determining biophysical parameters explaining biomolecular condensate formation, dissolution, and phase transitions. In addition, we mention how machine learning is increasingly important for efficient image analysis, teaching programs what a condensate should resemble, aiding in the correlation and interpretation of information from diverse data sources. Influenza A virus forms liquid viral inclusions in the infected cell cytosol that serve as model biomolecular condensates for this study. Our previous work showcased the possibility of hardening these liquid inclusions, potentially leading to novel antiviral strategies. This was established using a framework involving live cell imaging to measure dynamics, internal rearrangement capacity, coalescence, and relaxation time. Additionally, we integrated thermodynamic characteristics by analysing fixed images through Z-projections. The aforementioned paper laid the foundation for this subsequent technical paper, which explores how different modalities in data acquisition and processing impact the robustness of results to detect