Jian Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jian Liu has a total of 111 co-authors. Here are the top co-authors by number of shared publications:
- Jinmao Wei
College of Computer Science, Nankai University, Tianjin, China
9 shared publications Zhihan Ruan
Centre for Bioinformatics and Intelligent Medicine College of Computer Science Nankai University Tianjin 300350 China
8 shared publicationsJialiang Sun
College of Computer Science, Nankai University, Tianjin, China
6 shared publicationsWenting Xiang
Centre for Bioinformatics and Intelligent Medicine College of Computer Science Nankai University Tianjin 300350 China
6 shared publications- Chang Sun
College of Computer Science, Nankai University, Tianjin, China
5 shared publications - Chaoyang Yan
Centre for Bioinformatics and Intelligent Medicine College of Computer Science Nankai University Tianjin 300350 China
5 shared publications
Latest Publications
- Open Access
- 19 May 2026
- Published by Springer Nature in BMC Musculoskeletal Disorders
- Vol. 27 (1)
Role and mechanism of leptin in improving osteoporosis via the “gut–bone axis”
Osteoporosis is a systemic skeletal disorder characterized by reduced bone mass, microarchitectural deterioration, and increased fracture risk, particularl
- 19 May 2026
- Published by Springer Nature in BMC Musculoskeletal Disorders
- Vol. 27 (1)
Open AccessRole and mechanism of leptin in improving osteoporosis via the “gut–bone axis”
Osteoporosis is a systemic skeletal disorder characterized by reduced bone mass, microarchitectural deterioration, and increased fracture risk, particularl
- Open Access
- 12 February 2026
- Published by Elsevier in Cell Reports Methods
- Vol. 6 (2) , 101304
A unified framework combining linear and 3D molecular features for robust drug-protein interaction prediction
- 12 February 2026
- Published by Elsevier in Cell Reports Methods
- Vol. 6 (2) , 101304
Open AccessA unified framework combining linear and 3D molecular features for robust drug-protein interaction prediction
- Top Cited Papers
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (3) , 1754-1767
Feature Selection via Dynamic Feature Graph
In high-dimensional data, features often exhibit complex correlations and redundancies that hinder effective learning and reduce model interpretability. Therefore, extracting critical feature structural information from such data is essential for effective feature selection. This structural information can be represented as a feature graph, which reveals associations among features, including redundancies and correlations. However, most existing feature graph-based methods adopt a single strategy, focusing either on selecting relevant features or removing redundant ones, often resulting in suboptimal performance. Moreover, these methods typically construct static feature graphs using all input features, which may introduce irrelevant or detrimental relationships. Such simplistic approaches and low-quality graphs limit the effectiveness of feature selection. To address these limitations, we propose the Dynamic Feature Graph (DFG) framework that jointly learns feature graphs and selects features. By decoupling feature associations according to category, the DFG framework dynamically identifies key feature relationships, eliminating redundancy across classes while preserving essential intra-class correlations. This results in the selection of both relevant and non-redundant features. Additionally, we introduce sample manifolds with a rank equality constraint to ensure the selection of category-related features. To the best of our knowledge, this is the first dynamic feature graph approach in the field. Extensive experiments on 15 widely used real-world datasets demonstrate that DFG outperforms 20 state-of-the-art feature selection methods, highlighting its effectiveness in supervised and unsupervised settings.
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (3) , 1754-1767
Top Cited PapersFeature Selection via Dynamic Feature Graph
In high-dimensional data, features often exhibit complex correlations and redundancies that hinder effective learning and reduce model interpretability. Therefore, extracting critical feature structural information from such data is essential for effective feature selection. This structural information can be represented as a feature graph, which reveals associations among features, including redundancies and correlations. However, most existing feature graph-based methods adopt a single strategy, focusing either on selecting relevant features or removing redundant ones, often resulting in suboptimal performance. Moreover, these methods typically construct static feature graphs using all input features, which may introduce irrelevant or detrimental relationships. Such simplistic approaches and low-quality graphs limit the effectiveness of feature selection. To address these limitations, we propose the Dynamic Feature Graph (DFG) framework that jointly learns feature graphs and selects features. By decoupling feature associations according to category, the DFG framework dynamically identifies key feature relationships, eliminating redundancy across classes while preserving essential intra-class correlations. This results in the selection of both relevant and non-redundant features. Additionally, we introduce sample manifolds with a rank equality constraint to ensure the selection of category-related features. To the best of our knowledge, this is the first dynamic feature graph approach in the field. Extensive experiments on 15 widely used real-world datasets demonstrate that DFG outperforms 20 state-of-the-art feature selection methods, highlighting its effectiveness in supervised and unsupervised settings.
- Open Access
- 7 November 2025
- Published by Wiley in Advanced Science
- Vol. 13 (4) , e10307
Predicting MammaPrint Recurrence Risk from Breast Cancer Pathological Images Using a Weakly Supervised Transformer
Recurrence related to poor prognosis is a leading cause of mortality in patients with breast cancer (BC). The MammaPrint (MP) genomic assay is designed to stratify recurrence risk and evaluate chemotherapy benefits for early‐stage HR+/HER2‐ BC patients. However, MP fails to reveal spatial tumor morphology and is limited by high costs. In this study, a BC MP cohort is established and CPMP is developed, a weakly supervised agent‐attention transformer model, to predict MP recurrence risk from annotation‐free BC histopathological slides. CPMP achieves an AUROC of 0.824 ± 0.03 in predicting MP risk groups. CPMP is further leveraged for spatial and morphological analyses to explore histological patterns associated with MP risk groups. The model reveals tumor spatial localization at the whole‐slide level and highlights distinct intercellular interaction patterns of MP groups. It also characterizes the diversity in tumor morphology and uncovers MP high‐specific, low‐specific, and colocalized morphological phenotypes that differ in quantitative cellular composition. Prognostic evaluation in the external cohort exhibits significant stratification of distant metastasis risk (HR: 3.14, p ‐value = 0.0014), underscoring the prognostic power of CPMP. These findings demonstrate the capability of CPMP in MP risk prediction, offering a flexible supplement to genomic risk assessment in early‐stage BC.
- 7 November 2025
- Published by Wiley in Advanced Science
- Vol. 13 (4) , e10307
Open AccessPredicting MammaPrint Recurrence Risk from Breast Cancer Pathological Images Using a Weakly Supervised Transformer
Recurrence related to poor prognosis is a leading cause of mortality in patients with breast cancer (BC). The MammaPrint (MP) genomic assay is designed to stratify recurrence risk and evaluate chemotherapy benefits for early‐stage HR+/HER2‐ BC patients. However, MP fails to reveal spatial tumor morphology and is limited by high costs. In this study, a BC MP cohort is established and CPMP is developed, a weakly supervised agent‐attention transformer model, to predict MP recurrence risk from annotation‐free BC histopathological slides. CPMP achieves an AUROC of 0.824 ± 0.03 in predicting MP risk groups. CPMP is further leveraged for spatial and morphological analyses to explore histological patterns associated with MP risk groups. The model reveals tumor spatial localization at the whole‐slide level and highlights distinct intercellular interaction patterns of MP groups. It also characterizes the diversity in tumor morphology and uncovers MP high‐specific, low‐specific, and colocalized morphological phenotypes that differ in quantitative cellular composition. Prognostic evaluation in the external cohort exhibits significant stratification of distant metastasis risk (HR: 3.14, p ‐value = 0.0014), underscoring the prognostic power of CPMP. These findings demonstrate the capability of CPMP in MP risk prediction, offering a flexible supplement to genomic risk assessment in early‐stage BC.
- Open Access
- 1 November 2025
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (6)
A machine learning-based method to optimize the immunogenicity of human leukocyte antigen class I-restricted neoantigens
Neoantigens, arising from somatic mutations, have the potential to induce immune responses against tumor cells. As natural neoantigens that drive immune responses are uncommon, some methods to optimize neoantigens have been devised, leading to “heteroclitic” neoantigens capable of triggering cross-immunization. Existing methods, however, concentrate exclusively on the affinity of neoantigens for human leukocyte antigen (HLA), while neglecting the enhancement of their immunogenicity. Here, we developed a machine learning-based method, Naso, which integrates simulated annealing search and multi-objective training to optimize the immunogenicity of neoantigens. We also designed a search space optimization strategy to improve search efficiency. Experimental evaluation demonstrated that Naso outperforms existing methods in enhancing the immunogenicity of neoantigens. Specifically, Naso can transform nonimmunoreactive neoantigens into immunoreactive ones with no more than three mutation amino acids. Moreover, Naso can obtain competitive results in other immunological features, such as binding affinity and presentation. Consequently, Naso-optimized neoantigens can elicit an immune response and facilitate cross-immunization, thereby promoting the development of heteroclitic neoantigen-based vaccines. The source code of Naso is available at https://github.com/lyotvincent/Naso.
- 1 November 2025
- Published by Oxford University Press (OUP) in Briefings in Bioinformatics
- Vol. 26 (6)
Open AccessA machine learning-based method to optimize the immunogenicity of human leukocyte antigen class I-restricted neoantigens
Neoantigens, arising from somatic mutations, have the potential to induce immune responses against tumor cells. As natural neoantigens that drive immune responses are uncommon, some methods to optimize neoantigens have been devised, leading to “heteroclitic” neoantigens capable of triggering cross-immunization. Existing methods, however, concentrate exclusively on the affinity of neoantigens for human leukocyte antigen (HLA), while neglecting the enhancement of their immunogenicity. Here, we developed a machine learning-based method, Naso, which integrates simulated annealing search and multi-objective training to optimize the immunogenicity of neoantigens. We also designed a search space optimization strategy to improve search efficiency. Experimental evaluation demonstrated that Naso outperforms existing methods in enhancing the immunogenicity of neoantigens. Specifically, Naso can transform nonimmunoreactive neoantigens into immunoreactive ones with no more than three mutation amino acids. Moreover, Naso can obtain competitive results in other immunological features, such as binding affinity and presentation. Consequently, Naso-optimized neoantigens can elicit an immune response and facilitate cross-immunization, thereby promoting the development of heteroclitic neoantigen-based vaccines. The source code of Naso is available at https://github.com/lyotvincent/Naso.
- Open Access
Genomic and genetic insights into speciation and pigment pattern diversification in Danio fishes
The Danioninae subfamily of teleost fishes boasts up to four hundred distinct species that have evolved to display a stunning diversity of morphological forms. Here we use newly assembled genome sequences of four laboratory and wild zebrafish strains as well as eleven species of the Danio and Danionella genera to explore their phylogenetic history and the genetic basis of pigment pattern diversification. Phylogenomic analyses uncover extensive introgression and incomplete lineage sorting that have obscured phylogenetic relationships within Danio and corroborate an ancient hybrid origin of zebrafish. Whereas D. rerio inherited ancestral horizontal stripes, relatives repeatedly evolved spots and vertical bars. Interspecific complementation tests reveal functional divergence of the adhesion molecule gene igsf11 and the gap junction gene gja5b between the striped zebrafish and Danio species with divergent patterns. Comparative genomic and transcriptomic analyses suggest that protein and regulatory evolution have accompanied pigment pattern diversification. Our analyses elucidate complex genetic changes underlying the phylogenetic history and morphological diversification in the Danio genus. Resolved phylogenetic relationships, available genome assemblies, transcriptomes, and genetic tractability establish Danio fish species as excellent models for biomedical research in vertebrates.
Open AccessGenomic and genetic insights into speciation and pigment pattern diversification in Danio fishes
The Danioninae subfamily of teleost fishes boasts up to four hundred distinct species that have evolved to display a stunning diversity of morphological forms. Here we use newly assembled genome sequences of four laboratory and wild zebrafish strains as well as eleven species of the Danio and Danionella genera to explore their phylogenetic history and the genetic basis of pigment pattern diversification. Phylogenomic analyses uncover extensive introgression and incomplete lineage sorting that have obscured phylogenetic relationships within Danio and corroborate an ancient hybrid origin of zebrafish. Whereas D. rerio inherited ancestral horizontal stripes, relatives repeatedly evolved spots and vertical bars. Interspecific complementation tests reveal functional divergence of the adhesion molecule gene igsf11 and the gap junction gene gja5b between the striped zebrafish and Danio species with divergent patterns. Comparative genomic and transcriptomic analyses suggest that protein and regulatory evolution have accompanied pigment pattern diversification. Our analyses elucidate complex genetic changes underlying the phylogenetic history and morphological diversification in the Danio genus. Resolved phylogenetic relationships, available genome assemblies, transcriptomes, and genetic tractability establish Danio fish species as excellent models for biomedical research in vertebrates.
- Open Access
Atlas of Breast Cancer in Chinese Young Women Revealed by Single-cell RNA and ATAC Sequencing
Young women with breast cancer (YBC, age⩽40) are particularly prevalent in Asian. YBCs usually show more aggressive pathology and poorer outcomes than non-young patients. However, YBCs are underrepresented in current BC risk models, with their tumor intrinsic subtypes and microenvironments lacking a systematic elucidation at the single-cell level, thereby limiting the young-specific therapies. We established a single-cell Chinese YBC landscape baseline, including 246,659 cells, by applying scRNA-seq and scATAC-seq on untreated patients. We developed a cross-modal feature selection algorithm to construct a young-intrinsic subtype classifier ‘BCYtype’, outperforming existing classifiers in pseudobulk, cellular, and external cohorts. Comparative analyses with non-young samples revealed a direct differentiation trajectory from mammary stem cells to mature luminal cells. Pseudotemporal analysis also demonstrated that tumor cells in younger patients undergo earlier carcinogenesis. Mechanistically, we found that CDH1 interacts with pTex and NKT cells, serving as a young-specific marker and a potential therapeutic target for HR+ young patients. The interaction between exhausted T cells and antigen-presenting cells revealed NKG2A as a candidate therapeutic target for triple-negative breast cancer in young patients.
Open AccessAtlas of Breast Cancer in Chinese Young Women Revealed by Single-cell RNA and ATAC Sequencing
Young women with breast cancer (YBC, age⩽40) are particularly prevalent in Asian. YBCs usually show more aggressive pathology and poorer outcomes than non-young patients. However, YBCs are underrepresented in current BC risk models, with their tumor intrinsic subtypes and microenvironments lacking a systematic elucidation at the single-cell level, thereby limiting the young-specific therapies. We established a single-cell Chinese YBC landscape baseline, including 246,659 cells, by applying scRNA-seq and scATAC-seq on untreated patients. We developed a cross-modal feature selection algorithm to construct a young-intrinsic subtype classifier ‘BCYtype’, outperforming existing classifiers in pseudobulk, cellular, and external cohorts. Comparative analyses with non-young samples revealed a direct differentiation trajectory from mammary stem cells to mature luminal cells. Pseudotemporal analysis also demonstrated that tumor cells in younger patients undergo earlier carcinogenesis. Mechanistically, we found that CDH1 interacts with pTex and NKT cells, serving as a young-specific marker and a potential therapeutic target for HR+ young patients. The interaction between exhausted T cells and antigen-presenting cells revealed NKG2A as a candidate therapeutic target for triple-negative breast cancer in young patients.
- Open Access
Single-Cell RNA and ATAC Sequencing Reveal Cellular Heterogeneity and Chromatin Accessibility Dynamics in Young Chinese Breast Cancer Patients across Pre- and Post-Neoadjuvant Therapy
Young breast cancer (YBC) patients (age ≤ 40) exhibit aggressive features and poor prognosis, while facing complex needs including breast-conserving surgery and fertility preservation. Neoadjuvant therapy (NAT) is important for tumor downstaging and breast-conserving surgery in YBC treatment, but the NAT response mechanisms in YBC patients remain unelucidated. Here, we analyzed pre- and post-NAT samples from Chinese YBC patients using scRNA-seq and scATAC-seq. We found that NAT reshaped cellular heterogeneity in the tumor microenvironment (TME), reducing epithelial and T cells while increasing endothelial cells and fibroblasts. Residual cancer cells showed enriched epithelial-mesenchymal transition (EMT) and stromal remodeling programs. NAT responders had fewer luminal-like cells and retained basal-like cells in an early hybrid EMT state, whereas non-responders maintained both populations with late hybrid EMT. We identified therapy-resistant genes and motifs (e.g., VTCN1, PROM1, MZB1, Fox family) and linked CXCL13 upregulation to poor NAT response and tumor-specific T-cell expansion. Cell–cell communication analysis revealed NAT reprograms the TME by suppressing VEGF and TNF signaling in basal cells, while residual luminal cells transmit EGFR and CD47–SIRPA signals.
Open AccessSingle-Cell RNA and ATAC Sequencing Reveal Cellular Heterogeneity and Chromatin Accessibility Dynamics in Young Chinese Breast Cancer Patients across Pre- and Post-Neoadjuvant Therapy
Young breast cancer (YBC) patients (age ≤ 40) exhibit aggressive features and poor prognosis, while facing complex needs including breast-conserving surgery and fertility preservation. Neoadjuvant therapy (NAT) is important for tumor downstaging and breast-conserving surgery in YBC treatment, but the NAT response mechanisms in YBC patients remain unelucidated. Here, we analyzed pre- and post-NAT samples from Chinese YBC patients using scRNA-seq and scATAC-seq. We found that NAT reshaped cellular heterogeneity in the tumor microenvironment (TME), reducing epithelial and T cells while increasing endothelial cells and fibroblasts. Residual cancer cells showed enriched epithelial-mesenchymal transition (EMT) and stromal remodeling programs. NAT responders had fewer luminal-like cells and retained basal-like cells in an early hybrid EMT state, whereas non-responders maintained both populations with late hybrid EMT. We identified therapy-resistant genes and motifs (e.g., VTCN1, PROM1, MZB1, Fox family) and linked CXCL13 upregulation to poor NAT response and tumor-specific T-cell expansion. Cell–cell communication analysis revealed NAT reprograms the TME by suppressing VEGF and TNF signaling in basal cells, while residual luminal cells transmit EGFR and CD47–SIRPA signals.
- 25 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (6) , 3213-3223
HiADN: Lightweight Resolution Enhancement of Hi-C Data Using High Information Attention Distillation Network
Due to limitations in experimental library preparation and in practical sequencing cost, currently available Hi-C data is often sparse, affecting the precise characterization of complex 3D chromatin structures. Providing efficaciously computational models to elevate the quality of sparse Hi-C sequencing data for restoring the fundamental traits of 3D chromatin is of substantial significance. Herein, we introduce HiADN, a deep learning-based approach to infer dense high-resolution matrices from sparse Hi-C matrices. In particular, we firstly design a specialized architecture HiFM to captures local spatial structures and the patterns of Hi-C data. Then, we develop large kernel convolutional decomposition and attention mechanisms to effectively explore global patterns across longer genomic distances. Using HiADN, it is possible to construct biologically significant regions at high-resolution (e.g., 10Kb) while only using the 1/100 of original sequencing reads. The experimental results demonstrate that the effect of in silico libraries forecasted by computational models using HiADN is commensurate with that of experimental libraries, surpassing the state-of-the-art (SOTA) models. We further validated the effectiveness of HiADN in reconstructing the three-dimensional spatial structure of chromosomes on the GM12878, K562, and CH12-LX cell line datasets.
- 25 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Biology and Bioinformatics
- Vol. 22 (6) , 3213-3223
HiADN: Lightweight Resolution Enhancement of Hi-C Data Using High Information Attention Distillation Network
Due to limitations in experimental library preparation and in practical sequencing cost, currently available Hi-C data is often sparse, affecting the precise characterization of complex 3D chromatin structures. Providing efficaciously computational models to elevate the quality of sparse Hi-C sequencing data for restoring the fundamental traits of 3D chromatin is of substantial significance. Herein, we introduce HiADN, a deep learning-based approach to infer dense high-resolution matrices from sparse Hi-C matrices. In particular, we firstly design a specialized architecture HiFM to captures local spatial structures and the patterns of Hi-C data. Then, we develop large kernel convolutional decomposition and attention mechanisms to effectively explore global patterns across longer genomic distances. Using HiADN, it is possible to construct biologically significant regions at high-resolution (e.g., 10Kb) while only using the 1/100 of original sequencing reads. The experimental results demonstrate that the effect of in silico libraries forecasted by computational models using HiADN is commensurate with that of experimental libraries, surpassing the state-of-the-art (SOTA) models. We further validated the effectiveness of HiADN in reconstructing the three-dimensional spatial structure of chromosomes on the GM12878, K562, and CH12-LX cell line datasets.