Lianyi Han
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Lianyi Han has a total of 189 co-authors. Here are the top co-authors by number of shared publications:
- Feng Zhu
Zhejiang University
6 shared publications - Jingjing Xia
Greater Bay Area Institute of Precision Medicine (Guangzhou), School of Life Sciences, Fudan University, China
5 shared publications - Jiucun Wang
Human Phenome Institute, Fudan University, China
4 shared publications Lei Feng
Greater Bay Area Institute of Precision Medicine (Guangzhou), School of Life Sciences, Fudan University, Guangzhou, China
4 shared publications- Jean Krutmann
Environmentally-Induced Skin Aging Group, IUF-Leibniz Research Institute for Environmental Medicine, Düsseldorf, Germany; Medical Faculty, Heinrich Heine University Düsseldorf, Düsseldorf, Germany
3 shared publications - Mingyuan Qin
Department of Dermatology, Huashan Hospital, Shanghai Institute of Dermatology, Fudan University, Shanghai, China
3 shared publications
Latest Publications
- Open Access
- 25 November 2025
- Published by Springer Nature in Communications Biology
- Vol. 8 (1) , 1668
MoleculeFormer is a GCN-transformer architecture for molecular property prediction
Artificial intelligence is increasingly important in drug discovery, particularly in molecular property prediction. Graph Neural Networks can model molecular structures as graphs, using structural data to predict molecular properties and biological activities effectively. However, molecular feature optimization and model integration remain challenges. To address these challenges, we propose MoleculeFormer, a multi-scale feature integration model based on Graph Convolutional Network-Transformer architecture. It uses independent Graph Convolutional Network and Transformer modules to extract features from atom and bond graphs while incorporating rotational equivariance constraints and prior molecular fingerprints. The model captures both local and global features and introduces 3D structural information with invariance to rotation and translation. Experiments on 28 datasets show robust performance across various drug discovery tasks, including efficacy/toxicity prediction, phenotype screening, and ADME evaluation. The integration of attention mechanisms enhances interpretability, and the model demonstrates strong noise resistance, establishing MoleculeFormer as an effective, generalizable solution for molecular prediction tasks.
- 25 November 2025
- Published by Springer Nature in Communications Biology
- Vol. 8 (1) , 1668
Open AccessMoleculeFormer is a GCN-transformer architecture for molecular property prediction
Artificial intelligence is increasingly important in drug discovery, particularly in molecular property prediction. Graph Neural Networks can model molecular structures as graphs, using structural data to predict molecular properties and biological activities effectively. However, molecular feature optimization and model integration remain challenges. To address these challenges, we propose MoleculeFormer, a multi-scale feature integration model based on Graph Convolutional Network-Transformer architecture. It uses independent Graph Convolutional Network and Transformer modules to extract features from atom and bond graphs while incorporating rotational equivariance constraints and prior molecular fingerprints. The model captures both local and global features and introduces 3D structural information with invariance to rotation and translation. Experiments on 28 datasets show robust performance across various drug discovery tasks, including efficacy/toxicity prediction, phenotype screening, and ADME evaluation. The integration of attention mechanisms enhances interpretability, and the model demonstrates strong noise resistance, establishing MoleculeFormer as an effective, generalizable solution for molecular prediction tasks.
- Open Access
- 11 November 2025
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 54 (D1) , D1556-D1568
PersADE: a database of personalized adverse drug events and their underlying molecular mechanisms
As a major burden on global healthcare systems, adverse drug events (ADEs) result in significant morbidity, mortality, and healthcare resource consumption. With the rapid advances in precision medicine, personalized ADEs and their molecular mechanisms are important components of drug repurposing and drug safety improvement. Thus, extensive studies have been conducted to collect valuable information on personalized ADEs, but no database has yet been available to provide such data. In this work, PersADE, a database aiming to provide personalized drug adverse events and their molecular mechanisms, was constructed. It integrated 4 061 772 personalized drug-ADE associations, 31 756 protein-ADE associations, and 108 677 drug-protein interactions, with a particular emphasis on off-target effects. The uniqueness of these data lies in (a) providing demographic characteristics, disease context and drug administration parameters associated with ADEs, enabling stratification of drug-ADE associations; (b) systematically integrating interactions among drugs, human proteins and ADEs, describing the mechanistic insights. Given the growing global focus on precision medicine, PersADE is highly anticipated to significantly impact studies on personalized ADEs and mechanistic explorations by providing researchers and clinicians with evidence-based tools. It is now freely accessible at: https://idrblab.org/PersADE
- 11 November 2025
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 54 (D1) , D1556-D1568
Open AccessPersADE: a database of personalized adverse drug events and their underlying molecular mechanisms
As a major burden on global healthcare systems, adverse drug events (ADEs) result in significant morbidity, mortality, and healthcare resource consumption. With the rapid advances in precision medicine, personalized ADEs and their molecular mechanisms are important components of drug repurposing and drug safety improvement. Thus, extensive studies have been conducted to collect valuable information on personalized ADEs, but no database has yet been available to provide such data. In this work, PersADE, a database aiming to provide personalized drug adverse events and their molecular mechanisms, was constructed. It integrated 4 061 772 personalized drug-ADE associations, 31 756 protein-ADE associations, and 108 677 drug-protein interactions, with a particular emphasis on off-target effects. The uniqueness of these data lies in (a) providing demographic characteristics, disease context and drug administration parameters associated with ADEs, enabling stratification of drug-ADE associations; (b) systematically integrating interactions among drugs, human proteins and ADEs, describing the mechanistic insights. Given the growing global focus on precision medicine, PersADE is highly anticipated to significantly impact studies on personalized ADEs and mechanistic explorations by providing researchers and clinicians with evidence-based tools. It is now freely accessible at: https://idrblab.org/PersADE
- Open Access
Integrative Multi-omics Deciphers the Assembly Rules of the Human Skin Microbiome and Highlights Lipid-Mediated Niche Partitioning
Understanding the ecological assembly mechanisms of microbial communities is critical for deciphering host-microbiota interactions, however, such processes remain poorly characterized in cutaneous ecosystems. By profiling a non-industrialized rural cohort, this study deciphers the assembly rules governing the human skin microbiome and reveals an evolutionarily conserved core microbiota structured by niche selection. Microbial α-diversity peaks when C. acnes and M. osloensis reach equilibrium, supporting the Intermediate Disturbance Hypothesis. Host lipids, rather than aqueous compounds, emerged as the primary drivers of microbial niche partitioning. We further demonstrate that C. acnes and Malassezia restricta function as core metabolic engineers, expressing diverse lipases that hydrolyze triglycerides (TGs) into free fatty acids (FFAs), thereby maintaining skin surface lipid homeostasis (∼85% TG + FFA). Notably, virtually all skin surface FFAs are microbially derived and function as “public goods”, enabling cross-feeding that supports auxotrophic members and fosters mutualistic interactions. Importantly, in vitro coculture experiments confirm that C. acnes establish a survival niche for M. restricta . Our findings support a paradigm shift from a taxonomic to a functional understanding of the skin ecosystem, underscoring microbial lipid metabolism as a central orchestrator of ecological stability and a promising target for therapeutic intervention. Abstract Figure: GRAPHIC ABSTRACT
Open AccessIntegrative Multi-omics Deciphers the Assembly Rules of the Human Skin Microbiome and Highlights Lipid-Mediated Niche Partitioning
Understanding the ecological assembly mechanisms of microbial communities is critical for deciphering host-microbiota interactions, however, such processes remain poorly characterized in cutaneous ecosystems. By profiling a non-industrialized rural cohort, this study deciphers the assembly rules governing the human skin microbiome and reveals an evolutionarily conserved core microbiota structured by niche selection. Microbial α-diversity peaks when C. acnes and M. osloensis reach equilibrium, supporting the Intermediate Disturbance Hypothesis. Host lipids, rather than aqueous compounds, emerged as the primary drivers of microbial niche partitioning. We further demonstrate that C. acnes and Malassezia restricta function as core metabolic engineers, expressing diverse lipases that hydrolyze triglycerides (TGs) into free fatty acids (FFAs), thereby maintaining skin surface lipid homeostasis (∼85% TG + FFA). Notably, virtually all skin surface FFAs are microbially derived and function as “public goods”, enabling cross-feeding that supports auxotrophic members and fosters mutualistic interactions. Importantly, in vitro coculture experiments confirm that C. acnes establish a survival niche for M. restricta . Our findings support a paradigm shift from a taxonomic to a functional understanding of the skin ecosystem, underscoring microbial lipid metabolism as a central orchestrator of ecological stability and a promising target for therapeutic intervention. Abstract Figure: GRAPHIC ABSTRACT
- Open Access
- 5 September 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 272, 109070
ZeroTuneBio NER: A three-stage framework for zero-shot and zero-tuning biomedical entity extraction using large language models and prompt engineering
- 5 September 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 272, 109070
Open AccessZeroTuneBio NER: A three-stage framework for zero-shot and zero-tuning biomedical entity extraction using large language models and prompt engineering
- Open Access
- 31 August 2025
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (5)
Microllm: a structured information extraction tool using large language models and named entity recognition in microbiology
Understanding the characteristics of individual microorganisms is crucial for deciphering microbial community structures and enabling effective system manipulation. Large-scale analysis in microbial bioinformatics is hindered by the scarcity of high-quality annotated datasets, the labor-intensive, and time-consuming process of manual review, and the limitations of current natural language processing (NLP) technologies in effectively capturing comprehensive information. To tackle these challenges, we developed a new tool MicroLLM, which integrates fine-tuned large language models (LLMs) and bidirectional encoder representations from Transformers (BERT) to enhance the extraction of microbial phenotypic information. MicroLLM converts unstructured text into structured JSON objects, making the data more accessible and organized. By leveraging the strengths of LLMs and named entity recognition (NER), it efficiently extracts complex, multirelational, and multi-entity information. This method establishes a strong foundation for constructing large-scale structured knowledge bases, streamlining the creation of scientific knowledge, and advancing our understanding of microbial ecosystems and their impacts on health.
- 31 August 2025
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (5)
Open AccessMicrollm: a structured information extraction tool using large language models and named entity recognition in microbiology
Understanding the characteristics of individual microorganisms is crucial for deciphering microbial community structures and enabling effective system manipulation. Large-scale analysis in microbial bioinformatics is hindered by the scarcity of high-quality annotated datasets, the labor-intensive, and time-consuming process of manual review, and the limitations of current natural language processing (NLP) technologies in effectively capturing comprehensive information. To tackle these challenges, we developed a new tool MicroLLM, which integrates fine-tuned large language models (LLMs) and bidirectional encoder representations from Transformers (BERT) to enhance the extraction of microbial phenotypic information. MicroLLM converts unstructured text into structured JSON objects, making the data more accessible and organized. By leveraging the strengths of LLMs and named entity recognition (NER), it efficiently extracts complex, multirelational, and multi-entity information. This method establishes a strong foundation for constructing large-scale structured knowledge bases, streamlining the creation of scientific knowledge, and advancing our understanding of microbial ecosystems and their impacts on health.
- Open Access
- 1 April 2025
- Published by KEAI Publishing in Intelligent Oncology
- Vol. 1 (2) , 89-104
AI dermatology: Reviewing the frontiers of skin cancer detection technologies
- 1 April 2025
- Published by KEAI Publishing in Intelligent Oncology
- Vol. 1 (2) , 89-104
Open AccessAI dermatology: Reviewing the frontiers of skin cancer detection technologies
- Open Access
- 12 March 2025
- Published by MDPI AG in Bioengineering
- Vol. 12 (3) , 282
Deep Multi-Modal Skin-Imaging-Based Information-Switching Network for Skin Lesion Recognition
The rising prevalence of skin lesions places a heavy burden on global health resources and necessitates an early and precise diagnosis for successful treatment. The diagnostic potential of recent multi-modal skin lesion detection algorithms is limited because they ignore dynamic interactions and information sharing across modalities at various feature scales. To address this, we propose a deep learning framework, Multi-Modal Skin-Imaging-based Information-Switching Network (MDSIS-Net), for end-to-end skin lesion recognition. MDSIS-Net extracts intra-modality features using transfer learning in a multi-scale fully shared convolutional neural network and introduces an innovative information-switching module. A cross-attention mechanism dynamically calibrates and integrates features across modalities to improve inter-modality associations and feature representation in this module. MDSIS-Net is tested on clinical disfiguring dermatosis data and the public Derm7pt melanoma dataset. A Visually Intelligent System for Image Analysis (VISIA) captures five modalities: spots, red marks, ultraviolet (UV) spots, porphyrins, and brown spots for disfiguring dermatosis. The model performs better than existing approaches with an mAP of 0.967, accuracy of 0.960, precision of 0.935, recall of 0.960, and f1-score of 0.947. Using clinical and dermoscopic pictures from the Derm7pt dataset, MDSIS-Net outperforms current benchmarks for melanoma, with an mAP of 0.877, accuracy of 0.907, precision of 0.911, recall of 0.815, and f1-score of 0.851. The model’s interpretability is proven by Grad-CAM heatmaps correlating with clinical diagnostic focus areas. In conclusion, our deep multi-modal information-switching model enhances skin lesion identification by capturing relationship features and fine-grained details across multi-modal images, improving both accuracy and interpretability. This work advances clinical decision making and lays a foundation for future developments in skin lesion diagnosis and treatment.
- 12 March 2025
- Published by MDPI AG in Bioengineering
- Vol. 12 (3) , 282
Open AccessDeep Multi-Modal Skin-Imaging-Based Information-Switching Network for Skin Lesion Recognition
The rising prevalence of skin lesions places a heavy burden on global health resources and necessitates an early and precise diagnosis for successful treatment. The diagnostic potential of recent multi-modal skin lesion detection algorithms is limited because they ignore dynamic interactions and information sharing across modalities at various feature scales. To address this, we propose a deep learning framework, Multi-Modal Skin-Imaging-based Information-Switching Network (MDSIS-Net), for end-to-end skin lesion recognition. MDSIS-Net extracts intra-modality features using transfer learning in a multi-scale fully shared convolutional neural network and introduces an innovative information-switching module. A cross-attention mechanism dynamically calibrates and integrates features across modalities to improve inter-modality associations and feature representation in this module. MDSIS-Net is tested on clinical disfiguring dermatosis data and the public Derm7pt melanoma dataset. A Visually Intelligent System for Image Analysis (VISIA) captures five modalities: spots, red marks, ultraviolet (UV) spots, porphyrins, and brown spots for disfiguring dermatosis. The model performs better than existing approaches with an mAP of 0.967, accuracy of 0.960, precision of 0.935, recall of 0.960, and f1-score of 0.947. Using clinical and dermoscopic pictures from the Derm7pt dataset, MDSIS-Net outperforms current benchmarks for melanoma, with an mAP of 0.877, accuracy of 0.907, precision of 0.911, recall of 0.815, and f1-score of 0.851. The model’s interpretability is proven by Grad-CAM heatmaps correlating with clinical diagnostic focus areas. In conclusion, our deep multi-modal information-switching model enhances skin lesion identification by capturing relationship features and fine-grained details across multi-modal images, improving both accuracy and interpretability. This work advances clinical decision making and lays a foundation for future developments in skin lesion diagnosis and treatment.
Validity, reliability and responsiveness of digital visual analogue scales for chronic spontaneous urticaria monitoring: A CRUSE ® mobile health study
Background: CRUSE® is an app that allows patients with chronic spontaneous urticaria (CSU) to monitor their daily disease activity through the use of visual analogue scales (VASs). We aimed to determine the concurrent validity, reliability, responsiveness and minimal important difference (MID) of CRUSE® VASs. Method s: We evaluated the properties of three daily VASs: VAS for how much patients were affected by their CSU (‘VAS urticaria’), VAS for the impact of urticaria on work/school productivity (‘VAS productivity’) and the VAS of EQ‐5D. Concurrent validity was assessed by measuring the association between each VAS and the Urticaria Activity Score (UAS). Intra‐rater reliability was determined based on the data of users providing multiple daily questionnaires within the same day. Test–retest reliability and responsiveness (ability to change), respectively, were tested in clinically stable and clinically unstable users. MIDs were determined using distribution‐based methods. Results: We included 5938 patients (67,380 days). Concurrent validity was high, with VAS urticaria being more strongly associated with the UAS score than the remaining VASs. Intra‐rater reliability was also high, with intraclass correlation coefficients (ICC) being above 0.950 for all VASs. Moderate‐high test–retest reliability and responsiveness were observed, with reliability ICC being highest for VAS EQ‐5D and responsiveness being highest for VAS urticaria. The MID for VAS urticaria was 17 (out of 100) units, compared to 15 units for VAS productivity and 11 units for VAS EQ‐5D. Conclusion: Daily VASs for CSU available in the CRUSE® app display high concurrent validity and intra‐rater reliability and moderate‐high test–retest reliability and responsiveness.
Validity, reliability and responsiveness of digital visual analogue scales for chronic spontaneous urticaria monitoring: A CRUSE ® mobile health study
Background: CRUSE® is an app that allows patients with chronic spontaneous urticaria (CSU) to monitor their daily disease activity through the use of visual analogue scales (VASs). We aimed to determine the concurrent validity, reliability, responsiveness and minimal important difference (MID) of CRUSE® VASs. Method s: We evaluated the properties of three daily VASs: VAS for how much patients were affected by their CSU (‘VAS urticaria’), VAS for the impact of urticaria on work/school productivity (‘VAS productivity’) and the VAS of EQ‐5D. Concurrent validity was assessed by measuring the association between each VAS and the Urticaria Activity Score (UAS). Intra‐rater reliability was determined based on the data of users providing multiple daily questionnaires within the same day. Test–retest reliability and responsiveness (ability to change), respectively, were tested in clinically stable and clinically unstable users. MIDs were determined using distribution‐based methods. Results: We included 5938 patients (67,380 days). Concurrent validity was high, with VAS urticaria being more strongly associated with the UAS score than the remaining VASs. Intra‐rater reliability was also high, with intraclass correlation coefficients (ICC) being above 0.950 for all VASs. Moderate‐high test–retest reliability and responsiveness were observed, with reliability ICC being highest for VAS EQ‐5D and responsiveness being highest for VAS urticaria. The MID for VAS urticaria was 17 (out of 100) units, compared to 15 units for VAS productivity and 11 units for VAS EQ‐5D. Conclusion: Daily VASs for CSU available in the CRUSE® app display high concurrent validity and intra‐rater reliability and moderate‐high test–retest reliability and responsiveness.
- 18 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (1) , 166-176
rU-Net, Multi-Scale Feature Fusion and Transfer Learning: Unlocking the Potential of Cuffless Blood Pressure Monitoring With PPG and ECG
This study introduces an innovative deep-learning model for cuffless blood pressure estimation using PPG and ECG signals, demonstrating state-of-the-art performance on the largest clean dataset, PulseDB. The rU-Net architecture, a fusion of U-Net and ResNet, enhances both generalization and feature extraction accuracy. Accurate multi-scale feature capture is facilitated by short-time Fourier transform (STFT) time-frequency distributions and multi-head attention mechanisms, allowing data-driven feature selection. The inclusion of demographic parameters as supervisory information further elevates performance. On the calibration-based dataset, our model excels, achieving outstanding accuracy (SBP MAE ± std: 4.49 ± 4.86 mmHg, DBP MAE ± std: 2.69 ± 3.10 mmHg), surpassing AAMI standards and earning a BHS Grade A rating. Addressing the challenge of calibration-free data, we propose a fine-tuning-based transfer learning approach. Remarkably, with only 10% data transfer, our model attains exceptional accuracy (SBP MAE ± std: 4.14 ± 5.01 mmHg, DBP MAE ± std: 2.48 ± 2.93 mmHg). This study sets the stage for the development of highly accurate and reliable wearable cuffless blood pressure monitoring devices.
- 18 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (1) , 166-176
rU-Net, Multi-Scale Feature Fusion and Transfer Learning: Unlocking the Potential of Cuffless Blood Pressure Monitoring With PPG and ECG
This study introduces an innovative deep-learning model for cuffless blood pressure estimation using PPG and ECG signals, demonstrating state-of-the-art performance on the largest clean dataset, PulseDB. The rU-Net architecture, a fusion of U-Net and ResNet, enhances both generalization and feature extraction accuracy. Accurate multi-scale feature capture is facilitated by short-time Fourier transform (STFT) time-frequency distributions and multi-head attention mechanisms, allowing data-driven feature selection. The inclusion of demographic parameters as supervisory information further elevates performance. On the calibration-based dataset, our model excels, achieving outstanding accuracy (SBP MAE ± std: 4.49 ± 4.86 mmHg, DBP MAE ± std: 2.69 ± 3.10 mmHg), surpassing AAMI standards and earning a BHS Grade A rating. Addressing the challenge of calibration-free data, we propose a fine-tuning-based transfer learning approach. Remarkably, with only 10% data transfer, our model attains exceptional accuracy (SBP MAE ± std: 4.14 ± 5.01 mmHg, DBP MAE ± std: 2.48 ± 2.93 mmHg). This study sets the stage for the development of highly accurate and reliable wearable cuffless blood pressure monitoring devices.
- Open Access
- 7 October 2024
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 53 (D1) , D1683-D1691
MolBiC: the cell-based landscape illustrating molecular bioactivities
The measurement of cell-based molecular bioactivity (CMB) is critical for almost every step of drug development. With the booming application of AI in biomedicine, it is essential to have the CMB data to promote the learning of cell-based patterns for guiding modern drug discovery, but no database providing such information has been constructed yet. In this study, we introduce MolBiC, a knowledge base designed to describe valuable data on molecular bioactivity measured within a cellular context. MolBiC features 550 093 experimentally validated CMBs, encompassing 321 086 molecules and 2666 targets across 988 cell lines. Our MolBiC database is unique in describing the valuable data of CMB, which meets the critical demands for CMB-based big data promoting the learning of cell-based molecular/pharmaceutical pattern in drug discovery and development. MolBiC is now freely accessible without any login requirement at: https://idrblab.org/MolBiC/.
- 7 October 2024
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 53 (D1) , D1683-D1691
Open AccessMolBiC: the cell-based landscape illustrating molecular bioactivities
The measurement of cell-based molecular bioactivity (CMB) is critical for almost every step of drug development. With the booming application of AI in biomedicine, it is essential to have the CMB data to promote the learning of cell-based patterns for guiding modern drug discovery, but no database providing such information has been constructed yet. In this study, we introduce MolBiC, a knowledge base designed to describe valuable data on molecular bioactivity measured within a cellular context. MolBiC features 550 093 experimentally validated CMBs, encompassing 321 086 molecules and 2666 targets across 988 cell lines. Our MolBiC database is unique in describing the valuable data of CMB, which meets the critical demands for CMB-based big data promoting the learning of cell-based molecular/pharmaceutical pattern in drug discovery and development. MolBiC is now freely accessible without any login requirement at: https://idrblab.org/MolBiC/.