Flow dichroism of DNA can be quantitatively predicted via coarse-grained molecular simulations.

We demonstrate the use of multiscale polymer modeling to quantitatively predict DNA linear dichroism (LD) in shear flow. LD is the difference in absorption of light polarized along two perpendicular axes and has long been applied to study biopolymer structure and drug-biopolymer interactions. As LD is orientation dependent, the sample must be aligned in order to measure a signal. Shear flow via a Couette cell can generate the required orientation; however, it is challenging to separate the LD due to changes in polymer conformation from specific interactions, e.g., drug-biopolymer. In this study, we have applied a combination of Brownian dynamics and equilibrium Monte Carlo simulations to accurately predict polymer alignment, and hence flow LD, at modest computational cost. As the optical and conformational contributions to the LD can be explicitly separated, our findings allow for enhanced quantitative interpretation of LD spectra through the use of an in silico model to capture conformational changes. Our model requires no fitting and only five input parameters: the DNA contour length, persistence length, optical factor, solvent quality, and relaxation time, all of which have been well characterized in prior literature. The method is sufficiently general to apply to a wide range of biopolymers beyond DNA, and our findings could help guide the search for new pharmaceutical drug targets via flow LD.