Huaiyuan Zhu
Huaiyuan Zhu has a total of 52 co-authors. Here are the top co-authors by number of shared publications:
Yi Cao
China Tobacco Jiangsu Industrial Co. Ltd., Nanjing, China
5 shared publicationsFan Cao
China Tobacco Jiangsu Industrial Co., Ltd. Nanjing Jiangsu China
4 shared publicationsJing Liu
China Tobacco Jiangsu Industrial Co. Ltd., Nanjing, China
3 shared publicationsYuan Zhang
China Tobacco Jiangsu Industrial Co., Ltd Nanjing China
3 shared publicationsChangjian Wu
China Tobacco Jiangsu Industrial Co. Ltd., Nanjing, China
2 shared publications- Hua Zhang
Zhengzhou Tobacco Research Institute of CNTC, Zhengzhou 450001, China
2 shared publications
Latest Publications
- 8 September 2025
- Published by Wiley in Flavour and Fragrance Journal
- Vol. 40 (6) , 946-956
Interpretable Message‐Passing Neural Networks for Molecular Odour Perception Prediction: From Structural Topology to Olfactory Characterisation
Molecular odour, as a crucial chemical property, serves as the foundation for flavour modulation in food product development, directly impacting applications ranging from precision design of food flavours and fragrance synthesis to environmental monitoring. However, the structure–activity relationship between molecular characteristics and olfactory perception remains incompletely understood, while traditional experimental approaches face challenges in large‐scale molecular odour characterisation. Recent advances have highlighted the growing importance of artificial intelligence algorithms integrated with chemical data processing techniques for molecular property prediction and rational interpretation of structure‐flavour correlations. This study develops a MPNN‐based fusion model that constructs a multi‐label binary classification framework by integrating molecular topological features with chemical feature maps. Through atomic‐level feature embedding and bond relationship propagation mechanisms, our architecture effectively captures cross‐scale structural patterns from a meticulously curated molecular odour dataset. The model demonstrates robust predictive performance across key food‐related odour dimensions including Odourless, Fruity, Offensive, Oily and Woody (F1‐score: > 0.73, AUC: > 0.82), while maintaining strong generalisation capabilities for low‐abundance labels (e.g., Alcoholic, n = 50: F1‐score = 0.6940, AUC: = 0.8612). The mechanistic analysis revealed that gradient backpropagation can elucidate quantitative associations between specific functional groups and odour perception, providing actionable insights for rational modification of flavour molecules. Our work innovatively integrates multi‐label classification with feature visualisation, overcoming the interpretability limitations inherent in conventional QSAR models for odour prediction. The developed computational framework provides an extensible technical pathway for intelligent fragrance development and odour digitisation engineering, establishing a novel algorithmic paradigm for next‐generation food sensory innovation.
- 8 September 2025
- Published by Wiley in Flavour and Fragrance Journal
- Vol. 40 (6) , 946-956
Interpretable Message‐Passing Neural Networks for Molecular Odour Perception Prediction: From Structural Topology to Olfactory Characterisation
Molecular odour, as a crucial chemical property, serves as the foundation for flavour modulation in food product development, directly impacting applications ranging from precision design of food flavours and fragrance synthesis to environmental monitoring. However, the structure–activity relationship between molecular characteristics and olfactory perception remains incompletely understood, while traditional experimental approaches face challenges in large‐scale molecular odour characterisation. Recent advances have highlighted the growing importance of artificial intelligence algorithms integrated with chemical data processing techniques for molecular property prediction and rational interpretation of structure‐flavour correlations. This study develops a MPNN‐based fusion model that constructs a multi‐label binary classification framework by integrating molecular topological features with chemical feature maps. Through atomic‐level feature embedding and bond relationship propagation mechanisms, our architecture effectively captures cross‐scale structural patterns from a meticulously curated molecular odour dataset. The model demonstrates robust predictive performance across key food‐related odour dimensions including Odourless, Fruity, Offensive, Oily and Woody (F1‐score: > 0.73, AUC: > 0.82), while maintaining strong generalisation capabilities for low‐abundance labels (e.g., Alcoholic, n = 50: F1‐score = 0.6940, AUC: = 0.8612). The mechanistic analysis revealed that gradient backpropagation can elucidate quantitative associations between specific functional groups and odour perception, providing actionable insights for rational modification of flavour molecules. Our work innovatively integrates multi‐label classification with feature visualisation, overcoming the interpretability limitations inherent in conventional QSAR models for odour prediction. The developed computational framework provides an extensible technical pathway for intelligent fragrance development and odour digitisation engineering, establishing a novel algorithmic paradigm for next‐generation food sensory innovation.
- Open Access
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in IET Nanobiotechnology
- Vol. 2025 (1)
Ultrafast and Deep Saliva Proteome Reveals the Dynamic of Human Saliva With Aging by Orbitrap Astral Mass Spectrometer
Saliva has already proven to be a prospective diagnostic bioresource for both early disease detection and timely intervention due to its easy accessibility, noninvasiveness, and reproducibility. However, the in‐depth identification of salivary proteins needs to be further improved. Until now, only 3427 proteins are included in the human salivary proteome (HSP), which is far from the millions of proteins that make up humans. Here, we set out to quantitatively map the HSP in rapid and in‐depth Orbitrap Astral mass spectrometer (MS) and coronal nanomagnetic bead–based proteomics workflow. Our study reported 5937 salivary proteins, which was about 73% more than that recorded in HSP. Moreover, we compared the differences between the young and aged salivary proteins. The predominant functions of the upregulated proteins in the young were related to motor proteins and cardiomyopathy, whereas those of the aged were primarily upregulated with oxidation reaction, as well as neurodegenerative disorders. It is the first study to carry out salivary proteomics using a fast and deep Orbitrap Astral MS and remarkably enlarged the number of proteins with HSP, furthermore, salivary proteomics was found to be characterized in the young and aged. With the rapidly advancing MS and proteomics technologies, we believe that salivary protein biomarkers will be more promising for clinical diagnosis and prognosis of human diseases in the future.
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in IET Nanobiotechnology
- Vol. 2025 (1)
Open AccessUltrafast and Deep Saliva Proteome Reveals the Dynamic of Human Saliva With Aging by Orbitrap Astral Mass Spectrometer
Saliva has already proven to be a prospective diagnostic bioresource for both early disease detection and timely intervention due to its easy accessibility, noninvasiveness, and reproducibility. However, the in‐depth identification of salivary proteins needs to be further improved. Until now, only 3427 proteins are included in the human salivary proteome (HSP), which is far from the millions of proteins that make up humans. Here, we set out to quantitatively map the HSP in rapid and in‐depth Orbitrap Astral mass spectrometer (MS) and coronal nanomagnetic bead–based proteomics workflow. Our study reported 5937 salivary proteins, which was about 73% more than that recorded in HSP. Moreover, we compared the differences between the young and aged salivary proteins. The predominant functions of the upregulated proteins in the young were related to motor proteins and cardiomyopathy, whereas those of the aged were primarily upregulated with oxidation reaction, as well as neurodegenerative disorders. It is the first study to carry out salivary proteomics using a fast and deep Orbitrap Astral MS and remarkably enlarged the number of proteins with HSP, furthermore, salivary proteomics was found to be characterized in the young and aged. With the rapidly advancing MS and proteomics technologies, we believe that salivary protein biomarkers will be more promising for clinical diagnosis and prognosis of human diseases in the future.
- Open Access
- 13 December 2024
- Published by Wiley in Proteomics – Clinical Applications
- Vol. 19 (1) , e202400031
Comparisons of Whole Saliva and Cell Free Saliva by DIA‐Based Proteome Profiling
Background: Saliva has emerged as a promising diagnostic resource due to its accessibility, noninvasiveness, and repeatability, enabling early disease detection and timely intervention. However, current studies often overlook the distinction between whole saliva (WS) and cell‐free saliva (CFS).Objective: This study aims to compare the proteomic profiles of WS and CFS.Method and result: The saliva was detected with and without low‐abundance protein enrichment using nanoparticles, employing DIA‐MS technology. Our findings reveal a substantial enhancement in the detectability of low‐abundance proteins in saliva with utilization of nanoparticles, enabling identification of 12%–15% low‐abundance proteins previously undetectable in WS or CFS. In total, 3817 saliva proteins were identified, with 3413 found in WS and 2340 in CFS. More interestingly, we found that it was not the similarity of the samples that did the clustering, but rather it depended more on the different detection methods and sample types. And the predominant functions of the identified proteins in WS were related to oxidative phosphorylation and neurodegenerations, whereas those in CFS were primarily associated with nitrogen and glycosaminoglycan metabolism. And both exhibited functions in immune response and proteasome.Conclusion: This study represents the first comparison of WS and CFS, providing valuable experimental evidence for guiding the selection of research subjects in future saliva omics studies.
- 13 December 2024
- Published by Wiley in Proteomics – Clinical Applications
- Vol. 19 (1) , e202400031
Open AccessComparisons of Whole Saliva and Cell Free Saliva by DIA‐Based Proteome Profiling
Background: Saliva has emerged as a promising diagnostic resource due to its accessibility, noninvasiveness, and repeatability, enabling early disease detection and timely intervention. However, current studies often overlook the distinction between whole saliva (WS) and cell‐free saliva (CFS).Objective: This study aims to compare the proteomic profiles of WS and CFS.Method and result: The saliva was detected with and without low‐abundance protein enrichment using nanoparticles, employing DIA‐MS technology. Our findings reveal a substantial enhancement in the detectability of low‐abundance proteins in saliva with utilization of nanoparticles, enabling identification of 12%–15% low‐abundance proteins previously undetectable in WS or CFS. In total, 3817 saliva proteins were identified, with 3413 found in WS and 2340 in CFS. More interestingly, we found that it was not the similarity of the samples that did the clustering, but rather it depended more on the different detection methods and sample types. And the predominant functions of the identified proteins in WS were related to oxidative phosphorylation and neurodegenerations, whereas those in CFS were primarily associated with nitrogen and glycosaminoglycan metabolism. And both exhibited functions in immune response and proteasome.Conclusion: This study represents the first comparison of WS and CFS, providing valuable experimental evidence for guiding the selection of research subjects in future saliva omics studies.
- Open Access
- 31 May 2024
- Published by Wiley
Comparisons of Whole Saliva and Cell Free Saliva by DIA-Based Proteome Profiling
Saliva has emerged as a promising diagnostic resource due to its accessibility, non-invasiveness, and repeatability, enabling early disease detection and timely intervention. However, current studies often overlook the distinction between whole saliva (WS) and cell-free saliva (CFS). This study aims to compare the proteomic profiles of WS and CFS, both with and without low-abundance proteins enrichment using nanoparticles, employing DIA-MS technology. Our findings reveal a substantial enhancement in the detectability of low-abundance proteins in saliva with utilization of nanoparticles, enabling identification of 12%-15% low-abundance proteins previously undetectable in WS or CFS. In total, 3817 saliva proteins were identified, with 3413 found in WS and 2340 in CFS. More interestingly, we found that it wasn’t the similarity of the samples that did the clustering, but rather it depended more on the different detection methods and sample types. And the predominant functions of the identified proteins in WS were related to oxidative phosphorylation and neurodegenerations, whereas those in CFS were primarily associated with nitrogen and glycosaminoglycan metabolism. And both exhibited functions in immune response and proteasome. This study represents the first comparison of WS and CFS, providing valuable experimental evidence for guiding the selection of research subjects in future salivomics studies.
- 31 May 2024
- Published by Wiley
Open AccessComparisons of Whole Saliva and Cell Free Saliva by DIA-Based Proteome Profiling
Saliva has emerged as a promising diagnostic resource due to its accessibility, non-invasiveness, and repeatability, enabling early disease detection and timely intervention. However, current studies often overlook the distinction between whole saliva (WS) and cell-free saliva (CFS). This study aims to compare the proteomic profiles of WS and CFS, both with and without low-abundance proteins enrichment using nanoparticles, employing DIA-MS technology. Our findings reveal a substantial enhancement in the detectability of low-abundance proteins in saliva with utilization of nanoparticles, enabling identification of 12%-15% low-abundance proteins previously undetectable in WS or CFS. In total, 3817 saliva proteins were identified, with 3413 found in WS and 2340 in CFS. More interestingly, we found that it wasn’t the similarity of the samples that did the clustering, but rather it depended more on the different detection methods and sample types. And the predominant functions of the identified proteins in WS were related to oxidative phosphorylation and neurodegenerations, whereas those in CFS were primarily associated with nitrogen and glycosaminoglycan metabolism. And both exhibited functions in immune response and proteasome. This study represents the first comparison of WS and CFS, providing valuable experimental evidence for guiding the selection of research subjects in future salivomics studies.
- Open Access
- 19 February 2024
- Published by Elsevier in Neurobiology of Disease
- Vol. 193, 106442
The potential roles of salivary biomarkers in neurodegenerative diseases
- 19 February 2024
- Published by Elsevier in Neurobiology of Disease
- Vol. 193, 106442
Open AccessThe potential roles of salivary biomarkers in neurodegenerative diseases
- Open Access
Preparation of Functionalized Zr-Based MOFs and MOFs/GO for Efficient Removal of 1,3-Butadiene from Cigarette Smoke
Removal of 1,3-butadiene from cigarette smoke plays an important role in human health and environmental protection. Herein, a series of UiO-66 X% containing different ratios of the -NH2 group was synthesized via the solvothermal method by using terephthalic acid (H2BDC) and 2-aminoterephthalic acid (NH2-BDC) as ligands. Using GO as support, a series of UiO-66-NH2/GO Y% were prepared by controlling the ratio of UiO-66-NH2 and GO. The effects of -NH2 and GO contents on the structure and composition of MOFs were investigated. Finally, the different -NH2 contents of UiO-66 X% and the different GO contents of UiO-66-NH2/GO Y% were applied in 1,3-butadiene removal from cigarette smoke. The results showed that UiO-66 X% with the higher contents of -NH2 showed a higher rate of 1,3-butadiene removal, and UiO-66-NH2/GO Y% with the GO contents of 5% showed the highest removal rate of about 33.85%, which was 25.54% higher than that of activated carbon. In addition, the saturation capacity of the adsorbent materials for 1,3-butadiene was as high as 210.01–239.54 mg/g, showing great potential in reducing harmful components in cigarette smoke and environmental protection.
Open AccessPreparation of Functionalized Zr-Based MOFs and MOFs/GO for Efficient Removal of 1,3-Butadiene from Cigarette Smoke
Removal of 1,3-butadiene from cigarette smoke plays an important role in human health and environmental protection. Herein, a series of UiO-66 X% containing different ratios of the -NH2 group was synthesized via the solvothermal method by using terephthalic acid (H2BDC) and 2-aminoterephthalic acid (NH2-BDC) as ligands. Using GO as support, a series of UiO-66-NH2/GO Y% were prepared by controlling the ratio of UiO-66-NH2 and GO. The effects of -NH2 and GO contents on the structure and composition of MOFs were investigated. Finally, the different -NH2 contents of UiO-66 X% and the different GO contents of UiO-66-NH2/GO Y% were applied in 1,3-butadiene removal from cigarette smoke. The results showed that UiO-66 X% with the higher contents of -NH2 showed a higher rate of 1,3-butadiene removal, and UiO-66-NH2/GO Y% with the GO contents of 5% showed the highest removal rate of about 33.85%, which was 25.54% higher than that of activated carbon. In addition, the saturation capacity of the adsorbent materials for 1,3-butadiene was as high as 210.01–239.54 mg/g, showing great potential in reducing harmful components in cigarette smoke and environmental protection.
- Open Access
- 20 October 2021
- Published by MDPI AG in Separations
- Vol. 8 (11) , 196
Selective Separation and Analysis of Catecholamines in Urine Based on Magnetic Solid Phase Extraction by Mercaptophenylboronic Acid Functionalized Fe3O4-NH2@Au Magnetic Nanoparticles Coupled with HPLC
A novel magnetic solid phase extraction based on mercaptophenylboronic acid (MPBA)-functionalized Fe3O4-NH2@Au nanomaterial (Fe3O4-NH2@Au-MPBA) was developed for selective separation and enrichment of catecholamines (including dopamine, norepinephrine and adrenaline). Fe3O4-NH2@Au-MPBA nanoparticles were achieved by self-assembly-anchoring MPBA molecules on the surface of Fe3O4-NH2@Au nanocomposites, which were synthesized via a facial ultrasonic auxiliary in situ reduction process. The interaction between cis-diol from catecholamines and boronic acid was reversible and could be flexibly controlled by adjusting pH value. The catecholamines could be quickly adsorbed by Fe3O4-NH2@Au-MPBA in weak alkaline solution (pH 8.0–9.0) and subsequently released in acid solution (pH 1.0–2.0). The process of adsorption and dissociation was very fast. Furthermore, the three catecholamines could be detected in urine from children by high performance liquid chromatography (HPLC) with electrochemical detector. Under optimal conditions, norepinephrine (NE), epinephrine (EP) and dopamine (DA) were separated very well from internal standard and exhibited a good linearity in the range of 2.5–500.0 ng mL−1, with correlation coefficients of r2 > 0.9907. Limits of detection (LOD) (signal to noise = 3) were 0.39, 0.27 and 0.60 ng mL−1 for NE, EP and DA, respectively. Recoveries for the spiked catecholamines were in the range of 85.4–105.2% with the relative standard deviation (RSD) < 11.5%.
- 20 October 2021
- Published by MDPI AG in Separations
- Vol. 8 (11) , 196
Open AccessSelective Separation and Analysis of Catecholamines in Urine Based on Magnetic Solid Phase Extraction by Mercaptophenylboronic Acid Functionalized Fe3O4-NH2@Au Magnetic Nanoparticles Coupled with HPLC
A novel magnetic solid phase extraction based on mercaptophenylboronic acid (MPBA)-functionalized Fe3O4-NH2@Au nanomaterial (Fe3O4-NH2@Au-MPBA) was developed for selective separation and enrichment of catecholamines (including dopamine, norepinephrine and adrenaline). Fe3O4-NH2@Au-MPBA nanoparticles were achieved by self-assembly-anchoring MPBA molecules on the surface of Fe3O4-NH2@Au nanocomposites, which were synthesized via a facial ultrasonic auxiliary in situ reduction process. The interaction between cis-diol from catecholamines and boronic acid was reversible and could be flexibly controlled by adjusting pH value. The catecholamines could be quickly adsorbed by Fe3O4-NH2@Au-MPBA in weak alkaline solution (pH 8.0–9.0) and subsequently released in acid solution (pH 1.0–2.0). The process of adsorption and dissociation was very fast. Furthermore, the three catecholamines could be detected in urine from children by high performance liquid chromatography (HPLC) with electrochemical detector. Under optimal conditions, norepinephrine (NE), epinephrine (EP) and dopamine (DA) were separated very well from internal standard and exhibited a good linearity in the range of 2.5–500.0 ng mL−1, with correlation coefficients of r2 > 0.9907. Limits of detection (LOD) (signal to noise = 3) were 0.39, 0.27 and 0.60 ng mL−1 for NE, EP and DA, respectively. Recoveries for the spiked catecholamines were in the range of 85.4–105.2% with the relative standard deviation (RSD) < 11.5%.
- 18 October 2021
- Published by Springer Nature in Chemical Papers
- Vol. 76 (2) , 1097-1106
Investigation of novel mercapto-functionalized nanofiber membrane for the adsorption of mercury in air
- 18 October 2021
- Published by Springer Nature in Chemical Papers
- Vol. 76 (2) , 1097-1106
Investigation of novel mercapto-functionalized nanofiber membrane for the adsorption of mercury in air