#40 · Structural bio & protein design

Sequence patterning and copy number predict condensate phase behavior and structure

The common thread across these three preprints is the use of physically grounded, sequence- or copy-number-level models to predict condensate behavior. A unified intermediate-resolution framework combining HyRes protein and iConNA nucleic acid models is applied to predict condensate saturation concentrations and phase behavior. A simulation study of 18 LAF-1 RGG sequence variants at fixed composition indicates that charge and hydropathy patterning govern IDR phase separation and whether condensates mix or demix. The ecDNA polymer model extends this logic to nucleic acid condensates, reporting copy-number scaling of 3D architecture and low-noise enhanced oncogene regulatory contacts. What is new is the explicit linkage of sequence patterning and copy number to predicted phase outcomes and structural organization, rather than descriptive analysis of individual condensates. The ecDNA study addresses polymer architecture rather than phase separation per se, so this thread is not shared uniformly across all three items.