Twitxy
Saved
@binzmit's profile picture
View on X

Bin Zhang

@binzmit

Associate professor of chemistry, MIT
East Cambridge, Cambridge zhanggroup.mit.edu Joined November 2017
287 Following 1.2K Followers 167 Posts

Latest posts 5

Bin Zhang @binzmit
New preprint: CV-SVGD combines Stein variational gradient descent with parallel restrained MD to explore higher-dimensional collective-variable spaces. In peptide tests, it recovered free-energy surfaces with less MD sampling than umbrella sampling. arxiv.org/abs/2610.27429
arxiv.org Collective-Variable Stein Variational Gradient Descent for Free... Enhanced-sampling methods enable molecular simulations to access configurational regions that are difficult to explore with conventional dynamics, but their efficiency often deteriorates as the...
Bin Zhang @binzmit
Our new paper introduces a molecular dynamics model for limited-valency sticker interactions in biomolecular condensates. Simulations show spacer interactions can promote condensation while disrupting sticker-mediated organization and robust recruitment. doi.org/10.1021/acs.jp…
pubs.acs.org Valency-Limited Molecular Dynamics Simulations of Stickers-and-Spacers Polymers Reveal a Trade-Off... Abstract. Biomolecular condensates formed by intrinsically disordered proteins (IDPs) are often described using stickers-and-spacers models, in which speci
Bin Zhang @binzmit
How open is “open” chromatin? Our new Nature Genetics paper uses nucleosome-resolution modeling to reveal condensed euchromatin domains with short, accessible regulatory regions at the surface. nature.com/articles/s4158…
www.nature.com Euchromatin forms condensed domains with short active regions on the surface Nature Genetics - Simulations integrating micro-C and imaging data provide a coherent view of chromatin organization from nucleosomes to clutches to domains, revealing that most regions form...
2 replies 27 reposts 138 likes 11K views View on X
Bin Zhang @binzmit
Pretraining implicit solvent, or coarse grained, models is difficult, limiting the transferability of the resulting ML force fields. However, Justin just figured out a novel method for pretraining using protein language models. More can be found at: arxiv.org/abs/2601.05388
arxiv.org Knowledge Distillation of a Protein Language Model Yields a... Implicit solvent models (ISMs) promise to deliver the accuracy of explicit solvent simulations at a fraction of the computational cost. However, despite decades of development, their accuracy has...