Debayani Chakraborty, Achinta Sannigrahi, Mahammed Kaif, Arijit Patra, Deep Thakkar and Rahul Roy
Dengue virus, like other viral pathogens such as SARS-CoV-2 and Zika, relies on membrane fusion to enter host cells. This study explores how the oxidized cholesterol metabolite 25-hydroxycholesterol (25-HC) inhibits dengue virus infection. We show that 25-HC disrupts the virus’s ability to fuse with host cell membranes, significantly reducing viral entry and infection. Additionally, 25-HC upregulates CH25H, an interferon-stimulated gene that further suppresses viral replication. Notably, 25-HC exhibits strong synergistic effects when combined with other antiviral agents targeting different pathways. These findings highlight the potential of targeting host lipid pathways to block viral fusion and control dengue virus infection.
Harsh Chhajer and Rahul Roy
Effective experimental design is critical for understanding complex systems, yet traditional model-based design of experiments (MBDoE) can be limited by computational demands and oversimplified assumptions. To overcome these challenges, we introduce PARSEC (PARameter SEnsitivity Clustering) — a novel MBDoE framework that uses parameter sensitivity clustering to identify the most informative experimental measurements. Combined with a fast Approximate Bayesian Computation approach, PARSEC enables rapid assessment and ranking of experimental designs. Testing on two distinct kinetic models, we demonstrate that PARSEC improves parameter estimation, accommodates experimental constraints, and reveals a strong link between sample size and the optimal number of clusters. This approach offers a powerful new strategy for designing efficient, data-rich experiments by integrating system dynamics and model structure.
Nathan Ronceray, Salim Bennani, Marianna Fanouria Mitsioni, Nicole Siegel, Maria J. Marcaida, Claudio Bruschini, Edoardo Charbon, Rahul Roy, Matteo Dal Peraro, Guillermo P. Acuna , Aleksandra Radenovic1
We demonstrate a high-throughput method for fluorescence lifetime imaging microscopy (FLIM) at the single-molecule level using a commercial time-gated single-photon camera. Unlike traditional FLIM, which is accurate but slow, our approach captures lifetime data for over 3,000 molecules simultaneously with high spatial resolution. It achieves good precision—within threefold of standard methods like TCSPC—while enabling real-time imaging of protein interactions and single-molecule FRET at up to 25 Hz. This method offers a powerful new tool for applications such as multi-target molecular imaging and biopolymer sequencing.