BPS2026 – A fragment-based atomistic simulation approach to identify key sequence motifs driving phase separation of intrinsically disordered proteins

Intrinsically disordered proteins (IDPs) are the main components of nearly all biomolecular condensates. As individual residues, arginine, and aromatic amino acids are known to play leading roles in the phase separation underlying condensate formation. However, intermolecular interactions driving phase separation likely involve sequence motifs instead of just individual residues. 1 NMR spectroscopy has identified such key motifs in a few IDPs, 2 and sequence-based methods are starting to predict these motifs. 1 Here, we introduce a fragment-based atomistic simulation approach to identify key motifs in IDPs that drive their phase separation. An IDP is broken into overlapping fragments, and potential phase separation of each fragment is simulated at various concentrations using the SpiDec method. 3,4 We test this approach on the CAPRIN1 C-terminal region (residues 607–709), for which NMR has identified three key motifs: G 624 YR 626 , G 638 YR 640 , and R 660 DYSGYQ 666 . 2 We cover this disordered region with 12 fragments, each comprising 16 residues with an 8-residue shift. In SpiDec simulations, motif-containing fragments exhibit phase-separation tendencies, as shown by condensation into a slab. Intermolecular contact maps further reveal residue-residue interactions that drive phase separation. These results demonstrate that our fragment-based atomistic simulation approach may represent a general strategy for identifying key motifs and dissecting the determinants for the phase separation of IDPs. 1 Zhou (2025). JACS Au , 5, 4361. 2 Kay, K. (2021). PNAS 118, e2104897118. 3 Mazarakos, P., and Zhou (2022). Front. Mol. Biosci. 9, 1021939. 4 Zhou, Z. (2024). Cell Rep. Phys. Sci. 5, 102218.