Quanquan Gu

This paper introduces cryoFM, a generative foundation model for cryo-EM densities that improves reconstruction and refinement by combining data-driven priors with dataset-specific constraints.

The paper presents a cryo-EM conformational heterogeneity challenge to evaluate and improve methods for analyzing molecular motions in heterogeneous samples.

PhyGCN improves node representations in hypergraphs using self-supervised learning and variable hyperedge sizes, showing effectiveness on chromatin interactions and polypharmacy side-effects.

Implanting a lightweight structural adapter into a pretrained protein language model, with iterative refinement at inference, raises sequence recovery for given folds to 55.65%/56.63% on CATH 4.2/4.3.

This paper introduces a probabilistic model to compare multi-species functional genomic data, showing it can identify genomic regions with distinct evolutionary patterns of DNA replication timing across primates.