I. Multidimensional Integrated Research on Molecular Mechanisms in Psychiatric Disorders

Our group focuses on schizophrenia, a complex brain disorder characterized by typical polygenic inheritance patterns. As a common major psychiatric disorder, schizophrenia manifests as disturbances across multiple mental processes, including thought, perception, self-experience, cognition, emotion, and behavior, with thought disorder considered its core symptom. The disease has a lifetime prevalence of approximately 7.2‰ and a heritability of about 80%. Building upon our previous genomic research achievements (Nature Genetics 2011, 2017, 2019, etc.), our group further integrates large-scale whole-genome sequencing and third-generation long-read sequencing technologies to systematically screen and analyze high-risk genetic variants associated with neuropsychiatric disorders. Meanwhile, by utilizing patient-derived brain organoids and tissue samples, we integrate multi-omics data—including genomics, epigenomics, single-cell and spatial transcriptomics, proteomics, and metabolomics—to construct a neurodevelopmental molecular regulatory network for schizophrenia. Combined with artificial intelligence technology, we reveal abnormal patterns of key molecular pathways, providing novel strategies for early diagnosis and targeted intervention.

II. Exploration of Neural Structure and Function in Psychiatric Disease Models

Our group has established a functional validation system centered on CRISPR/Cas9 gene editing, reprogramming of peripheral blood mononuclear cells (PBMCs) from psychiatric patients, and brain organoid models. Based on this platform, we have successfully constructed gene-edited brain organoids carrying rare high-risk variants in genes such as SETD1A, GRIN2A, and TRIO, as well as brain organoids derived from families of schizophrenia patients, to systematically investigate the effects of these variants on neurodevelopment, synapse formation, neural circuit construction, and electrophysiological activity.

Our team is currently developing high-density neural signal acquisition brain-computer interfaces suitable for brain organoid models to enable real-time, high-density electrophysiological monitoring of brain organoids. On this foundation, we are developing AI algorithms to decode neural activity features in early developmental stages of patient-derived organoids, thereby elucidating disease pathogenesis. This system holds promise as a replacement for traditional animal models, advancing drug screening and development for psychiatric disorders.

Furthermore, our group has constructed animal models carrying high-risk gene mutations for psychiatric disorders, including both mouse and non-human primate models. We analyze neuroanatomical changes across multiple dimensions, including brain magnetic resonance imaging, single-cell whole-brain projectome, and synaptic morphology. Combined with neuronal electrophysiological signals and appropriate behavioral paradigms, we assess neural function to establish a comprehensive psychiatric disease evaluation system.

Using fluorescence micro-optical sectioning tomography (fMOST) technology, our team performs morphological analysis of whole-brain single-cell projection patterns of different neuronal types in the mouse ventral tegmental area and macaque cortical neurons. This approach explores neuron-type-specific projection preferences and the coordinated projection features of individual neurons to multiple brain regions. We will establish analytical methods suitable for systematic analysis of neural circuit differences in animal models of psychiatric disorders.

III. Innovation in Interdisciplinary Research Technologies and Methods

Data Analysis Methods and Platform (SHEsis)

Our group has independently developed the SHEsis (SHEsisPlus.Bio-X.cn) genetic analysis algorithm suite, which significantly enhances the fine-mapping capability of susceptibility genes for complex diseases. This platform provides critical algorithmic support for genome-wide association analysis, long-fragment haplotype analysis, gene-gene interaction studies, and drug-protein interaction mapping (Cell Research 2005, 2009, 2011).

Ultra-High-Throughput Gene Synthesis Technology

Through independent innovation in the principles, techniques, and instrumentation of ultra-high-throughput de novo nucleic acid synthesis, our group has pioneered breakthroughs including 4.35 million parallel synthesis capacity, single-strand lengths reaching 600 bp, and >90% accuracy for rapid automated in vitro assembly of 10 kb genomic fragments, with successful technology commercialization. Building upon this, we are integrating inkjet printing, fluidic reaction, environmental control, machine vision, and software control modules to systematically construct a DNA storage-dedicated synthesizer. Based on virtual-well microarray chips and picoliter-level inkjet printing, we will achieve petabyte-level archival cold data writing via high-throughput DNA synthesis.

Spatial Single-Cell Omics Chip Technology

Based on our independently developed high-throughput DNA synthesis and writing technology based on 3D inkjet printing, we have built a DNA synthesis platform and achieved groundbreaking, original results in printing DNA array chips. This technology will support large-scale spatial single-cell omics studies and mesoscopic atlas mapping of pituitary adenomas, specific brain functional regions, and even the entire brain.

IV. Exploration of Molecular Mechanisms in Other Complex Diseases

Pituitary Adenoma

Pituitary adenoma is a common intracranial benign tumor, accounting for 10%–20% of symptomatic central nervous system tumors, with imaging studies revealing small pituitary lesions in ~10% of normal populations. Our group has made breakthroughs in understanding its molecular mechanisms: leading the discovery of susceptibility genes (Nature Genetics 2015), revealing that ~70% of Cushing's disease (ACTH-secreting adenoma) is driven by somatic mutations in USP8, USP48, and BRAF (Cell Research 2015, Nature Communications 2018), and generating the most comprehensive somatic mutation atlas across pituitary adenoma subtypes (Cell Research 2016), providing a theoretical foundation for precision diagnosis and treatment. Currently, we are employing long-read sequencing to explore the complete variation profiles of invasive pituitary tumors, constructing spatial transcriptomic atlases of normal pituitary, and developing pituitary organoid cohorts using gene editing and stem cell technologies.

Obstructive Sleep Apnea Syndrome (OSA)

OSA is a common and serious sleep and respiratory disorder. Through large-scale genome-wide association analysis of 20,590 Han Chinese samples (American Journal of Respiratory and Critical Care Medicine 2022), we identified OSA-associated loci: an intronic variant rs6455893 in PACRG (risk variant) and a missense mutation rs3746804 in SLC52A3 (a riboflavin transporter) that is protective. Additionally, 18 loci linked to OSA quantitative traits and objective sleep characteristics were discovered. Functional studies showed rs3746804 correlates with higher serum riboflavin levels in humans, with functional validation in a knock-in mouse model demonstrating that this variant reduces OSA-related risk by promoting riboflavin (vitamin B2) uptake.

Polycystic Ovary Syndrome (PCOS)

PCOS is the most common endocrine and metabolic disorder among reproductive-age women, characterized by hyperandrogenism, persistent anovulation, and polycystic ovarian morphology, severely affecting reproductive health and often comorbid with depression, anxiety, obesity, insulin resistance, and type 2 diabetes. Our group has pioneered the identification of 94 PCOS genetic risk loci (Nature Genetics 2011, 2012, 2025), expanding understanding of its genetic architecture. Through multidimensional analyses integrating phenomics (with special focus on mood disorders), functional genomics, evolutionary genetics, and drug action networks, we systematically map PCOS pathogenic pathways and regulatory networks, providing essential support for its molecular early warning and precision diagnosis and treatment. 

Overall Goal

Our research group is dedicated to using psychiatric disorders as a breakthrough point, deeply integrating genomics, neuroscience, bioinformatics, and cutting-edge biotechnology to systematically dissect the genetic basis and molecular networks of brain diseases. By constructing and utilizing organoids, animal models, and independently innovated interdisciplinary technology platforms, we aim to reveal the core mechanisms from gene variants to neural circuit dysfunction, ultimately providing novel strategies and targets for precision diagnosis, early prediction, and intervention of brain diseases.