Biomedical subjects

Hao Zhu

Publications and source records attributed to Hao Zhu.

7 recordsLinked to original sources

PubMed2026

ToxiVerse: chemical bioprofiling, toxicity data sharing and customizable predictive modeling.

MOTIVATION: Chemical toxicity assessment is critical for drug development and environmental safety. Computational models have emerged as a promising alternative to animal testing and now play a significant role in efficiently evaluating new chemicals. To address the urgent need for user-friendly machine learning tools in computational toxicology, we developed ToxiVerse, a public web-based platform. RESULTS: ToxiVerse provides automatic chemical bioprofiling, curated toxicity datasets, and a predictive modeling interface designed for researchers who lack programming expertise. The platform comprises three integrated modules: (i) Bioprofiler, which provides chemical descriptors by combining chemical-bioactivity data from PubChem assays with a machine learning-based data gap-filling procedure; (ii) Database, which hosts ∼50 000 curated chemicals covering diverse toxicity endpoints; and (iii) Cheminformatics, which enables dataset upload, chemical curation, and automatic generation of quantitative structure-activity relationship models for toxicity prediction. AVAILABILITY: The tool is accessible at www.toxiverse.com, and source code is available at https://github.com/zhu-research-group/toxiverse.

Quantitative Structure-Activity Relationship↗

PubMed2026

Integrating structure and experimental data annotations with computational modeling framework for predicting micro-nanoplastics toxicities.

The wide use of plastic materials leads to increased emissions of micro-nanoplastics (MNPs) into the environment, raising significant concerns about their impact on human health. Traditional experimental approaches for assessing MNPs toxicity are costly, time-consuming, and there are no experimental protocols that are universally acceptable. Computational modeling using machine learning (ML) approaches provides an efficient alternative to MNP toxicity assessment. However, most modeling studies of MNPs are limited due to the lack of high-quality data and there are few previous modeling studies considering complex structures of MNPs for model training. To address this challenge, we constructed three MNP datasets with popular toxicity endpoints from various resources and used nanostructure annotation techniques to create virtual MNPs (vMNPs) for all MNP structures. The MNP structures were digitalized from annotated vMNPs, and geometrical descriptors were calculated using the Delaunay Tessellation approach. Moreover, important experimental information, such as concentrations and cell lines, were transformed into extra training variables. Partial least squares regression (PLSR) models were built using both experimental and geometrical descriptors and validated through a leave-one-out cross validation procedure. The resulting models showed reasonable performance in predicting toxicity potentials of MNPs for the three endpoints in the present datasets. Moreover, an additional library of vMNPs with their predicted properties and bioactivities was constructed, directing further research of new MNPs. This study provides three novel ML models for MNPs by integrating geometrical and experimental descriptors, which have the potential to assess new MNPs for their toxicity. The modeling strategy developed in this study can be easily expanded to model other MNP toxicity endpoints and create promising new models for MNP toxicity assessments.

Data annotation↗

PubMed2026

Preoperative Carbohydrate Supplementation Reduces Thirst and Improves Patient Satisfaction Before Elective Cesarean Delivery: A Randomized Controlled Trial.

BACKGROUND & AIMS: Prolonged preoperative fasting is a major source of patient discomfort, particularly thirst, before elective cesarean delivery. This study aimed to evaluate whether preoperative carbohydrate (CHO) supplementation could alleviate these discomforts and improve patient satisfaction without compromising safety. METHODS: In this single-center randomized controlled trial, 262 women scheduled for elective cesarean delivery under neuraxial anesthesia were randomly allocated to either the CHO group (Group CHO, n = 131), which received 355 mL of an oral carbohydrate solution on the night before and the morning of surgery, or the control group (Group C, n = 131), which followed conventional fasting. The primary outcome was the thirst Numeric Rating Scale (NRS, 0-10) score measured immediately before surgery. Secondary outcomes included hunger NRS, satisfaction NRS, and maternal and neonatal safety parameters. RESULTS: Baseline characteristics were comparable between groups. Despite a longer preoperative fasting duration in Group CHO (9.25 ± 1.05 vs. 8.74 ± 0.97 h, P < 0.001), this group exhibited significantly lower thirst NRS scores (1.69 ± 0.90 vs. 4.02 ± 0.99, P < 0.001) and hunger NRS scores (1.25 ± 0.94 vs. 2.09 ± 0.82, P < 0.001), as well as higher satisfaction NRS scores (7.70 ± 0.69 vs. 5.69 ± 1.17, P < 0.001). Subgroup analyses confirmed consistent benefits of CHO supplementation across most patient characteristics. Further analyses suggested that the maximum effect on thirst reduction occurred at approximately 9.2 h of solid fasting; however, the interaction between fasting duration and treatment group was not statistically significant (P = 0.187). CONCLUSION: Preoperative carbohydrate supplementation effectively reduces thirst and hunger and improves patient satisfaction before elective cesarean delivery without increasing maternal or neonatal risk. The beneficial effects were consistent across varying fasting durations, with exploratory spline analyses suggested a potential peak effect around 9.2 h, though this was not statistically significant and should be interpreted cautiously. These findings support the incorporation of carbohydrate loading into enhanced recovery protocols. TRIAL REGISTRATION: China Clinical Trial Registry ChiCTR2500097956.

Humans↗

PubMed2026

Advancing cancer detection and treatment using longitudinal routine clinical data.

Cancer management remains fragmented across its continuum, from late-stage diagnosis and salvage therapies to non-personalized surveillance. Here, we present Oncoformer, a unified multimodal transformer model trained on the China Oncology Multimodal Prediction and Surveillance Study (COMPASS) cohort (3.67 million individuals, 17.7 million clinical visits) and validated on independent external cohorts, including the UK Biobank. Oncoformer integrates longitudinal electronic health records with chest X-ray imaging to address multiple clinical tasks: pan-cancer diagnosis (area under the receiver operating characteristic curve [AUROC] = 0.956), future cancer prediction up to 1 year before diagnosis (AUROC = 0.869), tumor stage inference (mean AUROC > 0.90), patient-specific treatment-response forecasting, and recurrence-free survival stratification across ten cancer types (all p < 0.01). Staging predictions were independently validated against postoperative pathological endpoints and shown to converge on core cancer genomic pathways. By translating routine clinical data into a dynamic view of cancer evolution, Oncoformer provides a framework for risk-informed cancer prediction and treatment stratification using routine clinical data.

Humans↗

PubMed2025

Hepatocyte-specific CLSTN3B ablation impairs lipid droplet maturation and alleviates diet-induced steatohepatitis in mice.

Excessive lipid accumulation in hepatocytes, a hallmark of metabolic dysfunction-associated steatotic liver disease (MASLD), can lead to progressive liver damage. Understanding the molecular mechanisms governing lipid storage in hepatocytes is essential for identifying therapeutic targets to halt MASLD progression. Here, we show a pivotal role for the protein calsyntenin 3β (CLSTN3B) in promoting lipid droplet (LD) maturation and lipid storage in hepatocytes. Previously characterized as an endoplasmic reticulum (ER)-LD contact protein that facilitates LD maturation in adipocytes, we now show that CLSTN3B expression is strongly induced in mouse hepatocytes by peroxisome proliferator-activated receptor gamma (PPARγ) in response to dietary caloric excess. Hepatocyte-specific deletion of CLSTN3B in mice significantly increases energy expenditure, reduces metabolic efficiency, and protects against diet-induced hepatic steatosis and fibrosis. Mechanistically, CLSTN3B deficiency causes reduced LD phospholipid coverage and increased lipase recruitment. This results in enhanced fatty acid oxidation driven by a futile cycle of lipolysis and re-esterification. Notably, human clinical data reveal a positive correlation between hepatic CLSTN3B expression and MASLD severity and progression, emphasizing its relevance to human disease. Together, our findings establish CLSTN3B as a key regulator of hepatocyte lipid storage and metabolic efficiency and highlight its potential as a therapeutic target in MASLD.

Journal Article↗

PubMed2025

Deletion of hepatic FXR leads to more severe MASH development in female mice.

BACKGROUND: The farnesoid X receptor (FXR) has been identified as a therapeutic target for metabolic dysfunction-associated steatohepatitis (MASH). FXR is the major homeostatic regulator of bile acids (BAs) with dysregulation of BAs and/or FXR implicated in the pathogenesis of MASH. Synthetic whole-body FXR agonists have been developed to treat MASH. Although beneficial for MASH treatment, these whole-body modulators contribute to unfavorable side effects such as pruritus and an elevation in low-density liporoteins, thereby highlighting the importance of tissue and cell-restricted modulation of FXR in the development of novel therapeutics for MASH to negate potential harmful off-target effects. METHODS: The objective of this study was to determine the tissue-specific role of FXR in MASH development using male and female wild-type (WT), liver FXR KO (FXRhep-/-), intestinal FXR KO (FXRint-/-), and whole body FXR KO (FXR KO) mice fed either a low-fat control diet (CTL) or a MASH "Fast Food" (FF) diet. RESULTS: The results showed, in females, hepatic, but not intestinal, deficiency of FXR was associated with severe liver injury, through increased ALT, ALP, and genes indicative of inflammation and fibrosis when comparing FXRhep-/- versus FXRint-/-. Regardless of sex, hepatic FXR deficiency triggered the activation of neuroinflammation and neurodegenerative canonical pathways. CONCLUSIONS: These data suggest that hepatic FXR is more critical in suppressing liver injury during MASH development in female mice. However, this same trend was not clear in the male cohorts, highlighting sex differences and potential roles for sexual dimorphism in MASH development.

Animals↗

PubMed2024

The phosphatidylserine targeting antibody bavituximab plus pembrolizumab in unresectable hepatocellular carcinoma: a phase 2 trial.

Immune checkpoint inhibitors targeting PD-1/L1 have modest efficacy in hepatocellular carcinoma as single agents. Targeting membranous phosphatidylserine may induce pro-inflammatory and -immune stimulating effects that enhance immunotherapy activity. This hypothesis was tested in a single-arm phase 2 trial evaluating frontline bavituximab, a phosphatidylserine targeting antibody, plus pembrolizumab (anti-PD-1) in patients with unresectable hepatocellular carcinoma (NCT03519997). The primary endpoint was investigator-assessed objective response rate among evaluable patients, and secondary end points included progression-free survival, incidence of adverse events, overall survival, and duration of response. Among 28 evaluable patients, the confirmed response rate was 32.1%, which met the pre-specified endpoint, and the median progression-free survival was 6.3 months (95% CI, 1.3-11.3 months). Treatment related-adverse events of any grade occurred in 45.7% of patients, with grade 3 or greater adverse events in 14.3% of patients. Adverse events of any cause were observed in 33 patients (94.3%), with grade 3 or greater adverse events in 11 patients (31.4%). Prespecified exploratory analyses of baseline tumor specimens showed that a depletion of B cells, and the presence of fibrotic tissue and expression of immune checkpoints in stroma was associated with tumor response. These results suggest that targeting phosphatidylserine may lead to synergistic effects with PD-1 blockade without increasing toxicity rates, and future studies on this therapeutic strategy may be guided by biomarkers characterizing the pre-treatment tumor microenvironment.

Humans↗