Spectral demixing enables reliable dual color pair correlation function analysis of viral and cellular proteins in live cells.

Pair correlation function (pCF) is a powerful approach that has gained increasing popularity in recent years for studying protein dynamics in living cells; however, its application remains limited by methodological constraints and the absence of standardized protocols. In this work, we optimize key aspects of pCF and dual color-pCF analysis, showing that reliable results can be achieved with shorter acquisition times, reducing motion artifacts. We demonstrate that spectral bleed-through as low as 1% produces spurious cross-correlations, underscoring the need for rigorous correction. To address this, we introduce a robust signal demixing strategy that removes false-positive correlations while preserving genuine molecular interactions. This advance expands the reliability of correlation analyses in complex live-cell environments. Applying this approach, we investigate the nucleo-cytoplasmic shuttling of the splicing factor RBM10 in living cells expressing dengue virus (DENV) NS5 polymerase under poly(I:C)-triggered innate immune response. Our findings reveal that RBM10 and NS5 not only interact but also co-shuttle between the nuclear and cytoplasmic compartments. Furthermore, RBM10 transport is differentially modulated by both NS5 and innate immune induction, reinforcing the role of DENV in altering host nuclear transport. Overall, this work establishes signal demixing as a key advance for live-cell correlation techniques and demonstrates its potential for uncovering complex host-virus interactions.