Minsheng Hao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Minsheng Hao has a total of 116 co-authors. Here are the top co-authors by number of shared publications:
- Xuegong Zhang
MOE Key Laboratory of Bioinformatics and Bioinformatics Division of BNRIST, Department of Automation, Tsinghua University , Beijing 100084
20 shared publications - Lei Wei
Ministry of Education Key Laboratory of Bioinformatics, Center for Synthetic and Systems Biology, Beijing National Research Center for Information Science and Technology, Department of Automation, Tsinghua University , Beijing 100084
11 shared publications Haiyang Bian
MOE Key Laboratory of Bioinformatics and Bioinformatics Division of BNRIST, Department of Automation, Tsinghua University, Beijing, China
7 shared publicationsYixin Chen
MOE Key Laboratory of Bioinformatics and Bioinformatics Division of BNRIST, Department of Automation, Tsinghua University, Beijing, China
7 shared publicationsSijie Chen
Tsinghua University
5 shared publicationsChen Li
4 shared publications
Latest Publications
- Open Access
- 14 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
A multi-modal diffusion model with dual-cross-attention for multi-omics data generation and translation
Single-cell multi-omics technologies offer unprecedented opportunities to decipher complex cellular mechanisms. To overcome experimental limitations in scale, cost, and coverage, powerful computational methods are essential for integrating diverse data modalities and generating high-fidelity in-silico data. In this paper, we present scDiffusion-X, a latent diffusion model for multi-omics data integration, generation, and translation. The core innovation is a Dual-Cross-Attention (DCA) module that adaptively captures intricate, hidden relationships between molecular modalities, offering a more flexible and interpretable approach than existing integration strategies. Extensive benchmarking experiments demonstrate that scDiffusion-X excels at generating realistic multi-omics data, preserving cellular heterogeneity and global data structures with excellent scalability. Beyond simulation, scDiffusion-X uniquely enables accurate modality translation, predicting one molecular modality from another with robust uncertainty quantification. Furthermore, we designed a gradient-based interpretation framework to transform the DCA module into a discovery tool, enabling inference of comprehensive cell-type-specific heterogeneous gene regulatory networks (GRNs). By integrating state-of-the-art generative modeling with biological interpretability, scDiffusion-X serves as a powerful tool for dissecting regulatory relationships, predicting perturbation responses, and accelerating discovery in single-cell multi-omics research.
- 14 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessA multi-modal diffusion model with dual-cross-attention for multi-omics data generation and translation
Single-cell multi-omics technologies offer unprecedented opportunities to decipher complex cellular mechanisms. To overcome experimental limitations in scale, cost, and coverage, powerful computational methods are essential for integrating diverse data modalities and generating high-fidelity in-silico data. In this paper, we present scDiffusion-X, a latent diffusion model for multi-omics data integration, generation, and translation. The core innovation is a Dual-Cross-Attention (DCA) module that adaptively captures intricate, hidden relationships between molecular modalities, offering a more flexible and interpretable approach than existing integration strategies. Extensive benchmarking experiments demonstrate that scDiffusion-X excels at generating realistic multi-omics data, preserving cellular heterogeneity and global data structures with excellent scalability. Beyond simulation, scDiffusion-X uniquely enables accurate modality translation, predicting one molecular modality from another with robust uncertainty quantification. Furthermore, we designed a gradient-based interpretation framework to transform the DCA module into a discovery tool, enabling inference of comprehensive cell-type-specific heterogeneous gene regulatory networks (GRNs). By integrating state-of-the-art generative modeling with biological interpretability, scDiffusion-X serves as a powerful tool for dissecting regulatory relationships, predicting perturbation responses, and accelerating discovery in single-cell multi-omics research.
- Open Access
How different AI models understand cells differently
AI single-cell foundation models (scFMs) are believed to be able to learn essential relations in cell transcriptomics with the attention modules in Transformer, but there is no method to reveal what they actually learned. We observed that different models may grasp different aspects of relations. To unravel the mystery, we propose scGeneLens, a framework for dissecting how scFMs perceive cells. We employed a sparse block attention to replace the original attention mechanism to concentrate attentions into a few dominant gene–gene relations, used attention propagation to trace how the relations propagate across Transformer layers, and used integrated gradients to disentangle the relative contributions of gene identity and expression in cell representations. We applied it to scFoundation and scGPT and show that they exhibit pluralistic perceptions of cells: scFoundation emphasizes relations among cell-type marker genes, resulting in stronger cell-type separability, whereas scGPT focuses more on genes involved in shared cellular pathways and core biological activities, leading to representations that generalize across conditions. The framework provides a unified lens for probing what scFMs learn about cells and offers actionable insights for the design of future cellular foundation models. Our code can be seen in https://anonymous.4open.science/r/scGeneLens-B771/ .
Open AccessHow different AI models understand cells differently
AI single-cell foundation models (scFMs) are believed to be able to learn essential relations in cell transcriptomics with the attention modules in Transformer, but there is no method to reveal what they actually learned. We observed that different models may grasp different aspects of relations. To unravel the mystery, we propose scGeneLens, a framework for dissecting how scFMs perceive cells. We employed a sparse block attention to replace the original attention mechanism to concentrate attentions into a few dominant gene–gene relations, used attention propagation to trace how the relations propagate across Transformer layers, and used integrated gradients to disentangle the relative contributions of gene identity and expression in cell representations. We applied it to scFoundation and scGPT and show that they exhibit pluralistic perceptions of cells: scFoundation emphasizes relations among cell-type marker genes, resulting in stronger cell-type separability, whereas scGPT focuses more on genes involved in shared cellular pathways and core biological activities, leading to representations that generalize across conditions. The framework provides a unified lens for probing what scFMs learn about cells and offers actionable insights for the design of future cellular foundation models. Our code can be seen in https://anonymous.4open.science/r/scGeneLens-B771/ .
- Open Access
- 15 December 2025
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
hECA v2.0: an AI-ready ensemble cell atlas of single-cell RNA and ATAC sequencing data
With the growing accumulation of scattered single-cell data and the rapid advancement of artificial intelligence (AI), there is a pressing need for a high-quality, well-organized, and AI-ready single-cell data resources to support large-scale model. Here, we present version 2.0 of human Ensemble Cell Atlas (hECA), a cell atlas incorporating both single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC-seq) data. It expands the scRNA-seq data collection to 10,831,024 human cells with unified labels, and adds the new modality of scATAC-seq profiles with 1,450,511 cells. The data cover 42 human organs and tissues. To ensure cross-dataset consistency and quality, we standardized gene expression and chromatin accessibility matrices, harmonized cellular metadata, and manually re-annotated cell types based on the unified Hierarchical Annotation Framework (uHAF). The strength of the dataset has been shown in pre-training the large generative cellular AI model scMulan. hECA2.0 provides a well-structured and ready-to-use data resource, serving as a robust data foundation for AI-driven single-cell research.
- 15 December 2025
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
Open AccesshECA v2.0: an AI-ready ensemble cell atlas of single-cell RNA and ATAC sequencing data
With the growing accumulation of scattered single-cell data and the rapid advancement of artificial intelligence (AI), there is a pressing need for a high-quality, well-organized, and AI-ready single-cell data resources to support large-scale model. Here, we present version 2.0 of human Ensemble Cell Atlas (hECA), a cell atlas incorporating both single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC-seq) data. It expands the scRNA-seq data collection to 10,831,024 human cells with unified labels, and adds the new modality of scATAC-seq profiles with 1,450,511 cells. The data cover 42 human organs and tissues. To ensure cross-dataset consistency and quality, we standardized gene expression and chromatin accessibility matrices, harmonized cellular metadata, and manually re-annotated cell types based on the unified Hierarchical Annotation Framework (uHAF). The strength of the dataset has been shown in pre-training the large generative cellular AI model scMulan. hECA2.0 provides a well-structured and ready-to-use data resource, serving as a robust data foundation for AI-driven single-cell research.
- Open Access
BrainBridge Characterizes Key Factors affecting Alzheimer's Disease and Associated Phenotypes
Single-cell RNA sequencing (scRNA-seq) has significantly advanced our understanding of Alzheimer's disease and aging by revealing cellular heterogeneity and shifts in cell-type composition between diseased/old and healthy/young individuals. However, few existing studies utilize the rich information in single-cell transcriptomic atlases for robust patient-level modeling and biological feature selection. To address this gap, we present BrainBridge, a deep learning-based framework designed to integrate atlas-scale single-cell transcriptomic data with phenotypic information to model the biomolecular complexity of the human brain. BrainBridge functions both as a powerful predictor and an embedding model for representing sample-level expression profiles and covariates through comprehensive benchmarking. We also demonstrate its effectiveness in prioritizing key genes and cell types associated with disease progression, aging, and sex differences. We further validate our findings using external resources, including genome-wide and epigenome-wide association studies (GWAS and EWAS), spatial transcriptomics, and perturb-seq experiments. Finally, we deploy BrainBridge within an interactive, agent-powered interface that enables intuitive and user-friendly model interactions, promoting broader accessibility and application in biomedical research.
Open AccessBrainBridge Characterizes Key Factors affecting Alzheimer's Disease and Associated Phenotypes
Single-cell RNA sequencing (scRNA-seq) has significantly advanced our understanding of Alzheimer's disease and aging by revealing cellular heterogeneity and shifts in cell-type composition between diseased/old and healthy/young individuals. However, few existing studies utilize the rich information in single-cell transcriptomic atlases for robust patient-level modeling and biological feature selection. To address this gap, we present BrainBridge, a deep learning-based framework designed to integrate atlas-scale single-cell transcriptomic data with phenotypic information to model the biomolecular complexity of the human brain. BrainBridge functions both as a powerful predictor and an embedding model for representing sample-level expression profiles and covariates through comprehensive benchmarking. We also demonstrate its effectiveness in prioritizing key genes and cell types associated with disease progression, aging, and sex differences. We further validate our findings using external resources, including genome-wide and epigenome-wide association studies (GWAS and EWAS), spatial transcriptomics, and perturb-seq experiments. Finally, we deploy BrainBridge within an interactive, agent-powered interface that enables intuitive and user-friendly model interactions, promoting broader accessibility and application in biomedical research.
- Open Access
GeST: Towards Building A Generative Pretrained Transformer for Learning Cellular Spatial Context
Learning spatial context of cells through pretraining on spatial transcriptomics (ST) data may empower us to systematically decipher tissue organization and cellular interactions. Yet, transformer-based generative models often focus on modeling individual cells, neglecting the intricate spatial relationships within them. We develop GeST, a deep transformer model that is pretrained by a novel spatially informed generation task: Predict cellular expression profile of a given location based on the information from its neighboring cells. We propose a spatial attention mechanism for efficient pretraining, a flexible serialization strategy for converting ST data into sequences, and a cell tokenization method for quantizing gene expression profiles. We pretrained GeST on large-scale ST datasets of different ST technologies and demonstrated its superior performance in generating unseen spatial cells. Our results also show that GeST can extract spatial niche embeddings in a zero-shot way and can be further fine-tuned for spatial annotation tasks. Furthermore, GeST can simulate gene expression changes in response to perturbations of cells within spatial context, closely matching existing experimental results. Overall, GeST offers a powerful generative pre-training framework for learning spatial contexts in spatial transcriptomics.
Open AccessGeST: Towards Building A Generative Pretrained Transformer for Learning Cellular Spatial Context
Learning spatial context of cells through pretraining on spatial transcriptomics (ST) data may empower us to systematically decipher tissue organization and cellular interactions. Yet, transformer-based generative models often focus on modeling individual cells, neglecting the intricate spatial relationships within them. We develop GeST, a deep transformer model that is pretrained by a novel spatially informed generation task: Predict cellular expression profile of a given location based on the information from its neighboring cells. We propose a spatial attention mechanism for efficient pretraining, a flexible serialization strategy for converting ST data into sequences, and a cell tokenization method for quantizing gene expression profiles. We pretrained GeST on large-scale ST datasets of different ST technologies and demonstrated its superior performance in generating unseen spatial cells. Our results also show that GeST can extract spatial niche embeddings in a zero-shot way and can be further fine-tuned for spatial annotation tasks. Furthermore, GeST can simulate gene expression changes in response to perturbations of cells within spatial context, closely matching existing experimental results. Overall, GeST offers a powerful generative pre-training framework for learning spatial contexts in spatial transcriptomics.
- 10 April 2025
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 35 (6) , 1429-1439
TFcomb identifies transcription factor combinations for cellular reprogramming based on single-cell multiomics data
Reprogramming cell state transitions provides the potential for cell engineering and regenerative therapy. Finding the reprogramming transcription factors (TFs) and their combinations that can direct the desired state transition is crucial for the task. Computational methods have been developed to identify such reprogramming TFs. However, most of them can only generate a ranked list of individual TFs and ignore the identification of TF combinations. Even for individual reprogramming TF identification, current methods often fail to put the real effective reprogramming TFs at the top. To address these challenges, we developed TFcomb, a computational method that leverages single-cell multiomics data to identify reprogramming TFs and TF combinations. We modeled the task of finding reprogramming TFs and their combinations as an inverse problem, and used Tikhonov regularization to guarantee the generalization ability of solutions. For the coefficient matrix of the model, we designed a graph attention network to augment gene regulatory networks built with single-cell RNA-seq and ATAC-seq data. Benchmarking experiments on data of human embryonic stem cells demonstrate superior performance of TFcomb against existing methods for identifying individual TFs. We curate data sets of multiple cell reprogramming cases and demonstrate that TFcomb can efficiently identify reprogramming TF combinations from vast potential combinations. We apply TFcomb on a data set of mouse hair follicle development and find key TFs in cell differentiation. All experiments show that TFcomb is powerful in identifying reprogramming TFs and TF combinations from single-cell data sets to empower future cell engineering.
- 10 April 2025
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 35 (6) , 1429-1439
TFcomb identifies transcription factor combinations for cellular reprogramming based on single-cell multiomics data
Reprogramming cell state transitions provides the potential for cell engineering and regenerative therapy. Finding the reprogramming transcription factors (TFs) and their combinations that can direct the desired state transition is crucial for the task. Computational methods have been developed to identify such reprogramming TFs. However, most of them can only generate a ranked list of individual TFs and ignore the identification of TF combinations. Even for individual reprogramming TF identification, current methods often fail to put the real effective reprogramming TFs at the top. To address these challenges, we developed TFcomb, a computational method that leverages single-cell multiomics data to identify reprogramming TFs and TF combinations. We modeled the task of finding reprogramming TFs and their combinations as an inverse problem, and used Tikhonov regularization to guarantee the generalization ability of solutions. For the coefficient matrix of the model, we designed a graph attention network to augment gene regulatory networks built with single-cell RNA-seq and ATAC-seq data. Benchmarking experiments on data of human embryonic stem cells demonstrate superior performance of TFcomb against existing methods for identifying individual TFs. We curate data sets of multiple cell reprogramming cases and demonstrate that TFcomb can efficiently identify reprogramming TF combinations from vast potential combinations. We apply TFcomb on a data set of mouse hair follicle development and find key TFs in cell differentiation. All experiments show that TFcomb is powerful in identifying reprogramming TFs and TF combinations from single-cell data sets to empower future cell engineering.
- Top Cited PapersOpen Access
SpatialAgent: An autonomous AI agent for spatial biology
Advances in AI are transforming scientific discovery, yet spatial biology, a field that deciphers the molecular organization within tissues, remains constrained by labor-intensive workflows. Here, we present SpatialAgent, a fully autonomous AI agent dedicated for spatial-biology research. SpatialAgent integrates large language models with dynamic tool execution and adaptive reasoning. SpatialAgent spans the entire research pipeline, from experimental design to multimodal data analysis and hypothesis generation. Tested on multiple datasets comprising two million cells from human brain, heart, and a mouse colon colitis model, SpatialAgent’s performance surpassed the best computational methods, matched or outperformed human scientists across key tasks, and scaled across tissues and species. By combining autonomy with human collaboration, SpatialAgent establishes a new paradigm for AI-driven discovery in spatial biology.
Top Cited PapersOpen AccessSpatialAgent: An autonomous AI agent for spatial biology
Advances in AI are transforming scientific discovery, yet spatial biology, a field that deciphers the molecular organization within tissues, remains constrained by labor-intensive workflows. Here, we present SpatialAgent, a fully autonomous AI agent dedicated for spatial-biology research. SpatialAgent integrates large language models with dynamic tool execution and adaptive reasoning. SpatialAgent spans the entire research pipeline, from experimental design to multimodal data analysis and hypothesis generation. Tested on multiple datasets comprising two million cells from human brain, heart, and a mouse colon colitis model, SpatialAgent’s performance surpassed the best computational methods, matched or outperformed human scientists across key tasks, and scaled across tissues and species. By combining autonomy with human collaboration, SpatialAgent establishes a new paradigm for AI-driven discovery in spatial biology.
- Open Access
Multi-modal Diffusion Model with Dual-Cross-Attention for Multi-Omics Data Generation and Translation
Single-cell multi-omics technologies offer unprecedented opportunities to decipher complex cellular mechanisms. To overcome experimental limitations in scale, cost, and coverage, powerful computational methods are essential for integrating diverse data modalities and generating high-fidelity in-silico data. In this paper, we present scDiffusion-X, a latent diffusion model specifically designed for this purpose. The core innovation is a Dual-Cross-Attention (DCA) module that adaptively learns intricate, hidden relationships between different molecular modalities, offering a more flexible and interpretable approach than traditional integration strategies. Extensive benchmarking experiments demonstrate that scDiffusion-X excels at generating realistic multi-omics data, preserving cellular heterogeneity and global data structures with excellent scalability. Distinct from existing multi-omics simulators, scDiffusion-X uniquely enables high-fidelity modality translation by predicting one molecular modality from another and provides robust uncertainty quantification. Beyond data generation, we designed a gradient-based interpretation framework to transform DCA module into a discovery tool, enabling inference of comprehensive cell-type-specific heterogeneous gene regulatory networks (GRNs). By integrating state-of-the-art generative modeling with deep biological interpretability, scDiffusion-X serves as a powerful tool for dissecting regulatory relationships, predicting perturbation responses, poised to accelerate discovery in single-cell multi-omics research.
Open AccessMulti-modal Diffusion Model with Dual-Cross-Attention for Multi-Omics Data Generation and Translation
Single-cell multi-omics technologies offer unprecedented opportunities to decipher complex cellular mechanisms. To overcome experimental limitations in scale, cost, and coverage, powerful computational methods are essential for integrating diverse data modalities and generating high-fidelity in-silico data. In this paper, we present scDiffusion-X, a latent diffusion model specifically designed for this purpose. The core innovation is a Dual-Cross-Attention (DCA) module that adaptively learns intricate, hidden relationships between different molecular modalities, offering a more flexible and interpretable approach than traditional integration strategies. Extensive benchmarking experiments demonstrate that scDiffusion-X excels at generating realistic multi-omics data, preserving cellular heterogeneity and global data structures with excellent scalability. Distinct from existing multi-omics simulators, scDiffusion-X uniquely enables high-fidelity modality translation by predicting one molecular modality from another and provides robust uncertainty quantification. Beyond data generation, we designed a gradient-based interpretation framework to transform DCA module into a discovery tool, enabling inference of comprehensive cell-type-specific heterogeneous gene regulatory networks (GRNs). By integrating state-of-the-art generative modeling with deep biological interpretability, scDiffusion-X serves as a powerful tool for dissecting regulatory relationships, predicting perturbation responses, poised to accelerate discovery in single-cell multi-omics research.
- Open Access
- 25 September 2024
- Published by Oxford University Press (OUP) in National Science Review
- Vol. 11 (11) , nwae340
General-purpose pre-trained large cellular models for single-cell transcriptomics
Recent development of large language models (LLMs) in AI has inspired scientists to develop a few large-scale AI foundation models for single-cell transcriptomics or large cellular models (LCMs) pretrained on massive single-cell RNA-seq data. They illustrated superior performances on a wide spectrum of tasks although the models were only pretrained in a self-supervised manner, without any specific design for the downstream tasks. The success opened a promising new route toward using AI models to grasp the underlying biological knowledge from massive data to a scale that cannot be handled by human analysis. This review introduces the basic principles, major technical variations and typical applications of the current LCMs, and also shares our perspective on open questions and future directions of this exciting field.
- 25 September 2024
- Published by Oxford University Press (OUP) in National Science Review
- Vol. 11 (11) , nwae340
Open AccessGeneral-purpose pre-trained large cellular models for single-cell transcriptomics
Recent development of large language models (LLMs) in AI has inspired scientists to develop a few large-scale AI foundation models for single-cell transcriptomics or large cellular models (LCMs) pretrained on massive single-cell RNA-seq data. They illustrated superior performances on a wide spectrum of tasks although the models were only pretrained in a self-supervised manner, without any specific design for the downstream tasks. The success opened a promising new route toward using AI models to grasp the underlying biological knowledge from massive data to a scale that cannot be handled by human analysis. This review introduces the basic principles, major technical variations and typical applications of the current LCMs, and also shares our perspective on open questions and future directions of this exciting field.