Xuan Wang

TRU-PE improves prime editing efficiency and scalability for genome engineering in human cells, enabling multi-locus edits and a platform for studying genetic disorders.

This paper introduces scRegNet, a framework combining single-cell foundation models with graph-based learning to predict gene regulatory links, achieving state-of-the-art results on benchmark datasets.

TractEM provides manual labeling protocols for white matter tractography to improve reproducibility and accuracy in brain connectivity studies.

The paper introduces a multi-task learning framework for biomedical named entity recognition that improves performance by sharing information across entity types.