Daohua Jiang
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
Daohua Jiang has a total of 169 co-authors. Here are the top co-authors by number of shared publications:
- Bo Huang
College of Pharmaceutical Sciences, Capital Medical University
11 shared publications - Yan Zhao
National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
9 shared publications Di Wu
8 shared publicationsHuiwen Chen
Beijing National Laboratory for Condensed Matter Physics and Institute of Physics, Chinese Academy of Sciences, Beijing, China
8 shared publications- Zhuo Huang
State Key Laboratory of Natural and Biomimetic Drugs Department of Molecular and Cellular Pharmacology School of Pharmaceutical Sciences Peking University Beijing 100191 China
8 shared publications - Di Wu
Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing, China
7 shared publications
Latest Publications
- Open Access
- 26 October 2026
- Published by Elsevier in Cell Reports
- Vol. 45 (11)
Mechanisms of vesicular monoamine transporter 2 modulation by amphetamines and lobeline
- 26 October 2026
- Published by Elsevier in Cell Reports
- Vol. 45 (11)
Open AccessMechanisms of vesicular monoamine transporter 2 modulation by amphetamines and lobeline
- Open Access
- 18 August 2026
- Published by Springer Nature in Research Square
UniLingo3DMol: a unified 3D molecular language model for de novo and fragment-based drug design
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. Moreover, its 3D language model formulation also enables order-of-magnitude faster inference than existing generation models. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
- 18 August 2026
- Published by Springer Nature in Research Square
Open AccessUniLingo3DMol: a unified 3D molecular language model for de novo and fragment-based drug design
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. Moreover, its 3D language model formulation also enables order-of-magnitude faster inference than existing generation models. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
- Open Access
- 7 April 2026
- Published by Springer Nature in Cell Discovery
- Vol. 12 (1) , 26
Structural basis of glucosinolate recognition and transport by plant GTR1
Glucosinolates (GLSs) play crucial roles in plant defense against herbivores. GTR1 facilitates the high-affinity transport of GLSs through a proton-dependent process. However, the molecular mechanism underlying GLS recognition and transport by GTR1 remains largely unknown. Here, we present four cryo-EM structures of Arabidopsis GTR1 in distinct states, namely, the outward-apo, inward-apo, 4MTB-bound and 3IMG-bound forms, revealing the structural basis for GLS and proton cotransport by GTR1. GTR1 consists of an MFS-like transmembrane domain and an intracellular domain (ICD). The ICD plays an essential role in GTR1 function by interacting with the gating helix, transmembrane helix 7. GLSs are recognized by the central cavity residues and directly interact with the conserved E 1 X 1 X 2 E 2 K motif. Our structural and functional analyses demonstrated that the E 1 X 1 X 2 E 2 K motif and Glu513 determine the proton coupling of GTR1. This study provides mechanistic insights into how GTR1 transports GLSs, which could aid in improving crop quality and enhancing resistance to herbivory.
- 7 April 2026
- Published by Springer Nature in Cell Discovery
- Vol. 12 (1) , 26
Open AccessStructural basis of glucosinolate recognition and transport by plant GTR1
Glucosinolates (GLSs) play crucial roles in plant defense against herbivores. GTR1 facilitates the high-affinity transport of GLSs through a proton-dependent process. However, the molecular mechanism underlying GLS recognition and transport by GTR1 remains largely unknown. Here, we present four cryo-EM structures of Arabidopsis GTR1 in distinct states, namely, the outward-apo, inward-apo, 4MTB-bound and 3IMG-bound forms, revealing the structural basis for GLS and proton cotransport by GTR1. GTR1 consists of an MFS-like transmembrane domain and an intracellular domain (ICD). The ICD plays an essential role in GTR1 function by interacting with the gating helix, transmembrane helix 7. GLSs are recognized by the central cavity residues and directly interact with the conserved E 1 X 1 X 2 E 2 K motif. Our structural and functional analyses demonstrated that the E 1 X 1 X 2 E 2 K motif and Glu513 determine the proton coupling of GTR1. This study provides mechanistic insights into how GTR1 transports GLSs, which could aid in improving crop quality and enhancing resistance to herbivory.
- Open Access
- 27 February 2026
- Published by Springer Nature in Research Square
An SE(3)-equivariant language model for pocket-aware 3D molecular generation enables discovery of potent HPK1 inhibitors
Generating molecules that simultaneously achieve optimal 3D pocket-binding conformations and chemically plausible topologies remains a central challenge in AI for Structure-Based Drug Design. Graph-based models excel in SE(3)-equivariant spatial reasoning but often struggle to ensure chemical validity, whereas language models capture discrete chemical syntax yet lack 3D spatial understanding. Here we introduce SE3-BiLingoMol, an SE(3)-equivariant language model for pocket-aware 3D ligand de novo generation and fragment-guided optimization. Built upon Geometric Algebra Transformers and a fragment-aware SMILES representation, our model enables SE(3)-equivariant modeling of continuous 3D geometry while ensuring chemically valid molecular topologies. To counteract cumulative 3D conformational errors inherent to autoregressive generation, we developed a bidirectional attention-based self-refinement mechanism as a key architectural component of SE3-BiLingoMol. Our model achieves state-of-the-art performance in an in-silico evaluation across over 100 diverse targets. Critically, application of SE3-BiLingoMol led to the discovery of a novel tetracyclic HPK1 inhibitor showing potent in vitro activity and robust in vivo anti-tumor efficacy. This work demonstrates a powerful and practical generative AI framework for accelerating structure-based drug design.
- 27 February 2026
- Published by Springer Nature in Research Square
Open AccessAn SE(3)-equivariant language model for pocket-aware 3D molecular generation enables discovery of potent HPK1 inhibitors
Generating molecules that simultaneously achieve optimal 3D pocket-binding conformations and chemically plausible topologies remains a central challenge in AI for Structure-Based Drug Design. Graph-based models excel in SE(3)-equivariant spatial reasoning but often struggle to ensure chemical validity, whereas language models capture discrete chemical syntax yet lack 3D spatial understanding. Here we introduce SE3-BiLingoMol, an SE(3)-equivariant language model for pocket-aware 3D ligand de novo generation and fragment-guided optimization. Built upon Geometric Algebra Transformers and a fragment-aware SMILES representation, our model enables SE(3)-equivariant modeling of continuous 3D geometry while ensuring chemically valid molecular topologies. To counteract cumulative 3D conformational errors inherent to autoregressive generation, we developed a bidirectional attention-based self-refinement mechanism as a key architectural component of SE3-BiLingoMol. Our model achieves state-of-the-art performance in an in-silico evaluation across over 100 diverse targets. Critically, application of SE3-BiLingoMol led to the discovery of a novel tetracyclic HPK1 inhibitor showing potent in vitro activity and robust in vivo anti-tumor efficacy. This work demonstrates a powerful and practical generative AI framework for accelerating structure-based drug design.
- Open Access
- 17 February 2026
- Published by Springer Nature in Research Square
A unified language model bridging de novo and fragment-based 3D molecule design
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
- 17 February 2026
- Published by Springer Nature in Research Square
Open AccessA unified language model bridging de novo and fragment-based 3D molecule design
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
- 28 January 2026
- Published by Springer Nature in Nature
- Vol. 651 (8104) , 251-259
Structure and mechanism of the human bile acid transporter OSTα–OSTβ
Bile acids (BAs) are crucial amphipathic surfactants that function as multifaceted regulators in various physiological processes, including nutrient absorption and distribution, lipid metabolism and inflammation1,2. The human organic solute transporter αβ (OSTα–OSTβ; hereafter referred to as OSTα/β) is a BA transporter that has a key role in the secretion and distribution of BAs3,4,5,6. Pathogenic mutations in OSTα/β have been associated with cholestasis7,8. Despite the functional importance of OSTα/β in BA homeostasis, the stoichiometry and assembly of the complex and the molecular mechanism that underlies BA transport by OSTα/β remain unknown. Here we present cryo-electron microscopy structures of human OSTα/β in complex with cholesterols and an endogenous substrate, elucidating the structural basis for the function of OSTα/β. OSTα/β is assembled in a novel dimer-of-heterodimers manner: two OSTα units form the homodimeric core, with two OSTβ units bound to the periphery. OSTα adopts the G-protein-coupled-receptor (GPCR) fold and contains a unique cysteine-rich loop with seven palmitoylation sites; these cooperate with transmembrane helices 5 and 6, constituting a BA recognition site. A positive cavity in OSTα connects the BA site and facilitates the transmembrane translocation of BAs through OSTα/β. Together, this study reveals the architecture and transport mechanism of OSTα/β and provides insights into the structure–function relationships of this crucial transporter in BA homeostasis.
- 28 January 2026
- Published by Springer Nature in Nature
- Vol. 651 (8104) , 251-259
Structure and mechanism of the human bile acid transporter OSTα–OSTβ
Bile acids (BAs) are crucial amphipathic surfactants that function as multifaceted regulators in various physiological processes, including nutrient absorption and distribution, lipid metabolism and inflammation1,2. The human organic solute transporter αβ (OSTα–OSTβ; hereafter referred to as OSTα/β) is a BA transporter that has a key role in the secretion and distribution of BAs3,4,5,6. Pathogenic mutations in OSTα/β have been associated with cholestasis7,8. Despite the functional importance of OSTα/β in BA homeostasis, the stoichiometry and assembly of the complex and the molecular mechanism that underlies BA transport by OSTα/β remain unknown. Here we present cryo-electron microscopy structures of human OSTα/β in complex with cholesterols and an endogenous substrate, elucidating the structural basis for the function of OSTα/β. OSTα/β is assembled in a novel dimer-of-heterodimers manner: two OSTα units form the homodimeric core, with two OSTβ units bound to the periphery. OSTα adopts the G-protein-coupled-receptor (GPCR) fold and contains a unique cysteine-rich loop with seven palmitoylation sites; these cooperate with transmembrane helices 5 and 6, constituting a BA recognition site. A positive cavity in OSTα connects the BA site and facilitates the transmembrane translocation of BAs through OSTα/β. Together, this study reveals the architecture and transport mechanism of OSTα/β and provides insights into the structure–function relationships of this crucial transporter in BA homeostasis.
- Open Access
- 15 January 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Structural mechanisms for inhibition and activation of human small-conductance Ca2+-activated potassium channel SK2
The small-conductance calcium-activated potassium (SK1-3 or KCa2) channels regulate the intrinsic excitability and firing frequency of excitable cells. SK channels are modulated by a variety of distinct modulators; however, the underlying mechanisms remain elusive. Here, we present four cryoelectron microscopy structures of the human SK2-calmodulin complex bound with apamin, UCL1684, AP30663, and CAD-1883, elucidating their distinct binding sites and regulatory mechanisms. Apamin and UCL1684 compete for a similar binding site above the selectivity filter, which is formed by the distinct S3-S4 linker of SK2. CAD-1883 glues the N-lobe of calmodulin and the S4-S5 linker of SK2, reinforcing the open state. In contrast, AP30663 resides in the central cavity of SK2, blocking ion conductance. This study reveals multiple modulation sites in SK2 and the molecular mechanisms for the inhibition and potentiation of SK channels, which could advance rational drug design targeting SK2 channel for the treatment of cardiovascular and neurological disorders.
- 15 January 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessStructural mechanisms for inhibition and activation of human small-conductance Ca2+-activated potassium channel SK2
The small-conductance calcium-activated potassium (SK1-3 or KCa2) channels regulate the intrinsic excitability and firing frequency of excitable cells. SK channels are modulated by a variety of distinct modulators; however, the underlying mechanisms remain elusive. Here, we present four cryoelectron microscopy structures of the human SK2-calmodulin complex bound with apamin, UCL1684, AP30663, and CAD-1883, elucidating their distinct binding sites and regulatory mechanisms. Apamin and UCL1684 compete for a similar binding site above the selectivity filter, which is formed by the distinct S3-S4 linker of SK2. CAD-1883 glues the N-lobe of calmodulin and the S4-S5 linker of SK2, reinforcing the open state. In contrast, AP30663 resides in the central cavity of SK2, blocking ion conductance. This study reveals multiple modulation sites in SK2 and the molecular mechanisms for the inhibition and potentiation of SK channels, which could advance rational drug design targeting SK2 channel for the treatment of cardiovascular and neurological disorders.
- Open Access
A unified language model bridging de novo and fragment-based 3D molecule design delivers potent CBL-B inhibitors for cancer treatment
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
Open AccessA unified language model bridging de novo and fragment-based 3D molecule design delivers potent CBL-B inhibitors for cancer treatment
The rational design of small molecules is central to drug discovery, yet current artificial intelligence (AI) methodologies for generating three-dimensional (3D) molecules are often siloed, focusing on either de novo design or fragment-based design. The lack of a holistic framework limits AI’s application across the complex and multi-step pipeline spanning from novel scaffold identification to lead compound optimization, and prevents AI from effectively learning from the entire process. Here, we introduce UniLingo3DMol, a language model for 3D molecular generation, empowered by fragment permutation-capable molecular representation alongside multi-stage and multi-task training strategy. This integrated design enables UniLingo3DMol to seamlessly span both de novo and fragment-retained molecular design, demonstrating superior performance over existing generation models in in silico evaluations across more than 100 diverse biological targets. We further leveraged UniLingo3DMol in the design of inhibitors targeting CBL-B, a crucial immune E3 ubiquitin ligase and attractive immunotherapy target. This strategy led to a lead compound demonstrating excellent in vitro activity and robust in vivo anti-tumor efficacy. Our findings establish UniLingo3DMol as a generalized and powerful platform, showing the strong potential to advance AI-driven drug discovery.
- Open Access
- 24 October 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Structural basis for transport and inhibition of the human glucose-6-phosphate transporter G6PT
The human glucose-6-phosphate transporter (G6PT) moves glucose-6-phosphate (G6P) into the lumen of endoplasmic reticulum, playing a vital role in glucose homeostasis. Dysregulation of G6PT causes glycogen storage disease 1b. Despite its functional importance, the structure, G6P recognition, and inhibition mechanism of G6PT remain unclear. Here, we report the cryo-EM structures of human G6PT in apo, G6P-bound, and the specific inhibitor chlorogenic acid (CHA)-bound forms, elucidating the structural basis for G6PT transport and inhibition. The G6P pocket comprises subsite A for phosphate and subsite B for glucose. The CHA occupies the G6P site and locks G6PT in a partly-occluded state. Functional assays demonstrate that G6PT activity is enhanced by co-expression of glucose-6-phosphatase (G6PC), but G6PT does not form a complex with G6PC. Together, this study provides a solid foundation for understanding the structure‒function relationships and pathology of G6PT and sheds light on the future development of potential therapeutics targeting G6PT.
- 24 October 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Open AccessStructural basis for transport and inhibition of the human glucose-6-phosphate transporter G6PT
The human glucose-6-phosphate transporter (G6PT) moves glucose-6-phosphate (G6P) into the lumen of endoplasmic reticulum, playing a vital role in glucose homeostasis. Dysregulation of G6PT causes glycogen storage disease 1b. Despite its functional importance, the structure, G6P recognition, and inhibition mechanism of G6PT remain unclear. Here, we report the cryo-EM structures of human G6PT in apo, G6P-bound, and the specific inhibitor chlorogenic acid (CHA)-bound forms, elucidating the structural basis for G6PT transport and inhibition. The G6P pocket comprises subsite A for phosphate and subsite B for glucose. The CHA occupies the G6P site and locks G6PT in a partly-occluded state. Functional assays demonstrate that G6PT activity is enhanced by co-expression of glucose-6-phosphatase (G6PC), but G6PT does not form a complex with G6PC. Together, this study provides a solid foundation for understanding the structure‒function relationships and pathology of G6PT and sheds light on the future development of potential therapeutics targeting G6PT.