Y. Guan

The paper analyzes H5N1 influenza virus evolution in humans and ducks in Cambodia, finding human-adapting mutations but limited within-host evolution due to selection pressures.

This study develops quantitative metrics and simulated datasets to evaluate subclonal reconstruction methods in tumours.

This study uses deep residual neural networks to infer four-taxon phylogenies, outperforming existing methods under substitution heterogeneities.

The paper uses single-cell ATAC-seq and deep learning to analyze pancreatic islet cells, revealing cell-specific regulatory signatures linked to type 2 diabetes.

DeepSleep uses deep learning to automatically detect sleep arousals in polysomnographic data with high accuracy, resolution, and speed.

This study demonstrates that transient expression of Runx1 and Hoxa9 in pluripotent stem cells generates functional T cells capable of immune surveillance and tumor eradication.

This study identifies MnJ, a colloidal manganese salt, as a potent universal adjuvant that induces both humoral and cellular immune responses, including CTL activation and mucosal immunity.

The paper reports a 50% reduction in PE2,3,4 missing proteins on Chromosome 17 using protein-level evidence from proteomics data.

The paper presents a large-scale cancer pharmacogenomic screen with 11,576 experiments on 910 drug combinations across 85 cancer cell lines, enabling crowd-sourced prediction of synergistic drug pairs and biomarkers.

The paper develops a high-throughput method to analyze both T-cell-receptor-beta alleles in single cells using Next Generation Sequencing.

YAMDA accelerates EM-based motif discovery by using PyTorch and GPU, achieving thousandfold speedup while maintaining MEME-level accuracy.

The paper introduces a label-free optical imaging method to measure sub-nanometer mechanical motion during action potentials in single mammalian neurons, finding amplitude variations across neurons and regions.

The paper demonstrates that machine learning can accurately predict odor perception from molecular features, enabling reverse-engineering of smell.

This paper develops a probabilistic method to model functional relationships at the splice isoform level by integrating heterogeneous genomic data in mice, revealing isoform-specific interactions and functional diversity.