#38 · Structural bio & protein design

Simulations link sequence patterning and confinement to condensate organization

The dominant thread is simulation-based analysis of how sequence and organizational parameters govern condensate formation and structure. Two preprints address sequence-encoded phase behavior: one combines high-throughput molecular simulations, machine learning, and mixed-integer programming to engineer phase-separating peptide nanocondensates, and the other introduces shuffle-based normalization of SCD and SHD patterning metrics with a Monte Carlo design algorithm for disordered polypeptides with tunable phase behavior. The published item uses MD simulations to show that HBV pgRNA forms a hollow, shell-like LLPS condensate within the capsid, which aids polymerase mobility. The set is heterogeneous in system and method, so the shared thread is phase separation as a principle governed by sequence and confinement rather than a common method. What is new is the normalization of SCD and SHD metrics against shuffled controls, and the observation that pgRNA condenses into a hollow geometry under capsid confinement.