Histology–Genomics Integration

We develop statistical and deep-learning frameworks (HISNUC) that link histological image features (including nuclear morphology, tissue texture, and cellular organization) to genotype, transcriptome, and chronological age. This approach has identified novel imageQTLs and demonstrated that gene expression and age can be predicted directly from tissue morphology.

HISNUC framework, PNAS 2025
ImageQTLs HISNUC Whole-Slide Imaging Deep Learning
Meng et al., PNAS 2025

Neurogenomics and Gene Regulation

We investigate cell-type–specific gene regulation in the human brain, with a focus on psychiatric disorders and aging. As part of the PsychENCODE Consortium, we analyzed millions of nuclei from 388 human prefrontal cortex samples, uncovering molecular mechanisms of schizophrenia, autism, and bipolar disorder. We are also interested in how spatial organization reshapes gene regulation within each cortical layer, leveraging spatial transcriptomic data.

brainSCOPE Resource, Science 2024
snRNA-seq Regulatory Networks Psychiatric Disease Spatial Transcriptomics
Emani*, Liu*, …, Meng* et al., Science 2024

AI-Enabled Pathogen Genomics and Surveillance

We also develop AI-driven frameworks that connect health data with pathogen genomics. In partnership with the Wadsworth Center (New York State Department of Health), we enable scalable analysis of infection-associated tissue compartments and real-time outbreak surveillance.

Pathogen Genomics AMR
Meng*, Xing*, Chang* et al., Nat. Comm. 2024