Hui Zhang’s research while affiliated with Jinan University and other places

Background Gastric cancer (GC) remains a prevalent and lethal malignancy worldwide, underscoring the urgent need to identify novel therapeutic targets and elucidate the tumor microenvironment (TME) to enhance clinical outcomes. Methods IL36RN mRNA expression in GC tissues was analyzed using The Cancer Genome Atlas (TCGA) dataset. Bioinformatics approaches, cellular models, and clinical tissue microarrays were employed to investigate the functional role of IL36RN, its interactions within the TME, and its prognostic significance. Results IL36RN expression was markedly upregulated in GC tissues and associated with unfavorable survival outcomes. Functional assays demonstrated that IL36RN silencing suppressed GC cell proliferation and invasion. Elevated IL36RN expression correlated with enhanced CD8⁺ T cell infiltration in the TME and served as an independent prognostic indicator in GC. Conclusions IL36RN represents a potential prognostic biomarker and therapeutic target in GC, offering novel avenues for precision oncology and immunotherapeutic intervention.

Gastric cancers are one of the leading causes of cancer-related mortality worldwide. Characterized by high tumor heterogeneity, the treatment of gastric cancers remains challenging. Prior attempts to elucidate the tumor heterogeneity of gastric cancers have included subtyping it into differing clinical and molecular subtypes. For example, the main histological type, adenocarcinoma, has intestinal and diffuse subtypes. At the genomic level, gastric cancers are classified as either Epstein-Barr virus (EBV), microsatellite instability (MSI), chromosomal instability (CIN), or genomically stable (GS). Transcriptomic subtyping has further revealed subtype and clinical prognosis relationships: MSI, microsatellite-stable microsatellite stable with epithelial-to-mesenchymal transition phenotype (MSS/EMT) (poor prognosis), and MSS/TP53 active or inactive (intermediate prognosis). However, these subtypes are still poorly understood at the functional level. To address this gap, the Clinical Proteomics Tumor Analysis Consortium (CPTAC) conducted a comprehensive multi-omic characterization of 159 gastric tumors and 31 normal adjacent tissues. Moving beyond conventional genomic and transcriptomic analyses, we employed a novel multi-omics approach that encompasses the proteome, phosphoproteome, glycoproteome, metabolome, and additional “omes” such as ubiquitinome, acetylome, and protein-protein interactome. In addition, to better understand the unique role of the microbiota in gastric cancers, we integrated the entire microbiome genome and transcriptome data into the whole analysis. By combining these 12 layers of data and microbiome analysis together, our study provides unprecedented proteogenomic insights, advancing our understanding of the molecular mechanisms of gastric cancer oncogenesis and aiding in the development of novel precision oncologist therapies. In this study, we classified gastric cancers by molecular and clinical subtypes, revealing both shared and subtype-specific biological features at the multi-omic level. We highlight the associations of post-translational modification (PTM) alterations to immune and metabolic pathways. Proteogenomic analysis uncovered promising new surface targets and potential biomarkers for existing targeted therapies. Further in-depth analyses of PTMs and protein-protein interactions have uncovered insights into intracellular signaling pathways and PTM crosstalk with therapeutic potential. Additionally, the impact of microbiome on immune responses and PTM patterns underscores the interplay between microbiota and cancer phenotypes, providing valuable insights for advancing precision medicine in gastric cancers. Overall, our study provides the most comprehensive multi-omics atlas for gastric cancers, offering deep insights into its cancer biology and molecular heterogeneity. Citation Format Lindsey K. Olsen, Yuefan Wang, Hui Zhang, Bing Zhang, Clinical Proteomic Tumor Analysis Consortium (CPTAC). Comprehensive multi-omics characterization of gastric cancers with proteogenomic insights [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7474.

Diffusion models, as a type of generative model, have achieved impressive results in generating images and videos conditioned on textual conditions. However, the generation process of diffusion models involves denoising dozens of steps to produce photorealistic images/videos, which is computationally expensive. Unlike previous methods that design ``one-size-fits-all'' approaches for speed up, we argue denoising steps should be sample-specific conditioned on the richness of input texts. To this end, we introduce AdaDiff, a lightweight framework designed to learn instance-specific step usage policies, which are then used by the diffusion model for generation. AdaDiff is optimized using a policy gradient method to maximize a carefully designed reward function, balancing inference time and generation quality. We conduct experiments on three image generation and two video generation benchmarks and demonstrate that our approach achieves similar visual quality compared to the baseline using a fixed 50 denoising steps while reducing inference time by at least 33%, going as high as 40%. Furthermore, our method can be used on top of other acceleration methods to provide further speed benefits. Lastly, qualitative analysis shows that AdaDiff allocates more steps to more informative prompts and fewer steps to simpler prompts.

Non-motor symptoms are clinical manifestations of amyotrophic lateral sclerosis (ALS). However, few studies have examined these symptoms in patients with early-stage ALS. We conducted a cross-sectional study to explore non-motor symptoms in 69 patients with ALS within 18 months of disease onset. The Pittsburgh Sleep Quality Index (PSQI), the Epworth Sleepiness Scale (ESS), and the Hospital Anxiety and Depression Scale (HADS) were used to evaluate sleep quality, daytime sleepiness, and anxiety and depression, respectively. Differences in the abovementioned non-motor symptoms between ALS patients and age-/sex-matched caregivers were examined, and correlations between these symptoms and the clinical features of ALS were analyzed. Compared to caregivers, ALS patients were more likely to report poor sleep [odds ratio (OR) and 95% confidence interval (95% CI) = 2.664, 1.276–5.560; p = 0.009] and excessive daytime sleepiness (EDS) [OR and 95% CI = 5.135, 1.640–16.072; p = 0.005]. The PSQI scores in ALS patients correlated significantly with the disease progression rate [ΔFS = (48-score on the Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised, ALSFRS-R)/disease duration] [β(95% CI) = 2.867 (0.397, 5.336), p = 0.024] and plasma neurofilament light chain (NfL) levels [β (95% CI) = 0.041 (0.012, 0.070), p = 0.008). Our results revealed that the patients with early-stage ALS experienced poor sleep quality and daytime sleepiness and suggested that low sleep quality may be related to more rapid disease progression. Confounders were not obvious in the early stage of ALS, and our results suggested that these symptoms may be related to more severe and extensive pathological changes in the central nervous system.