Pub Date : 2026-01-05DOI: 10.1038/s43588-025-00934-2
Lulu Yan, Dongyan Zhang, Xiaoqiang Sun
In multicellular systems, cell fate determination emerges from the integration of intracellular signaling and intercellular communication. Spatial transcriptomics (ST) provides opportunities to elucidate these regulatory processes, yet inferring the spatiotemporal dynamics of cell state transitions (CSTs) governed by cell-cell communication (CCC) remains a challenge. Here we introduce CCCvelo to reconstruct CCC-driven CST dynamics by jointly optimizing a dynamic CCC signaling network and a latent CST clock. CCCvelo formulates a unified multiscale nonlinear kinetic model that integrates intercellular ligand-receptor signaling gradients with intracellular transcription-factor activation cascades to capture gene expression dynamics encoding CSTs. Moreover, we devise PINN-CELL, a physics-informed neural-network-based coevolution learning algorithm, which simultaneously optimizes model parameters and pseudotemporal ordering. Application of CCCvelo to high-resolution ST datasets, including mouse cortex, embryonic trunk development and human prostate cancer datasets, demonstrates its ability to successfully recover known morphogenetic trajectories while uncovering dynamic CCC signaling rewiring that orchestrates CST progression.
{"title":"Decoding cell state transitions driven by dynamic cell-cell communication in spatial transcriptomics.","authors":"Lulu Yan, Dongyan Zhang, Xiaoqiang Sun","doi":"10.1038/s43588-025-00934-2","DOIUrl":"https://doi.org/10.1038/s43588-025-00934-2","url":null,"abstract":"<p><p>In multicellular systems, cell fate determination emerges from the integration of intracellular signaling and intercellular communication. Spatial transcriptomics (ST) provides opportunities to elucidate these regulatory processes, yet inferring the spatiotemporal dynamics of cell state transitions (CSTs) governed by cell-cell communication (CCC) remains a challenge. Here we introduce CCCvelo to reconstruct CCC-driven CST dynamics by jointly optimizing a dynamic CCC signaling network and a latent CST clock. CCCvelo formulates a unified multiscale nonlinear kinetic model that integrates intercellular ligand-receptor signaling gradients with intracellular transcription-factor activation cascades to capture gene expression dynamics encoding CSTs. Moreover, we devise PINN-CELL, a physics-informed neural-network-based coevolution learning algorithm, which simultaneously optimizes model parameters and pseudotemporal ordering. Application of CCCvelo to high-resolution ST datasets, including mouse cortex, embryonic trunk development and human prostate cancer datasets, demonstrates its ability to successfully recover known morphogenetic trajectories while uncovering dynamic CCC signaling rewiring that orchestrates CST progression.</p>","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":" ","pages":""},"PeriodicalIF":18.3,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145907321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-02DOI: 10.1038/s43588-025-00919-1
Jeheon Woo, Seonghwan Kim, Jun Hyeong Kim, Woo Youn Kim
Here we introduce a framework for molecular structure optimization using a denoising model on a physics-informed Riemannian manifold (R-DM). Unlike conventional approaches operating in Euclidean space, our method leverages a Riemannian metric that better aligns with molecular energy change, enabling more robust modeling of potential energy surfaces. By incorporating internal coordinates reflective of energetic properties, R-DM achieves chemical accuracy with an energy error below 1 kcal mol-1. Comparative evaluations on QM9, QM7-X and GEOM datasets demonstrate improvements in both structural and energetic accuracy, surpassing conventional Euclidean-based denoising models. This approach highlights the potential of physics-informed coordinates for tackling complex molecular optimization problems, with implications for tasks in computational chemistry and materials science.
{"title":"Riemannian denoising model for molecular structure optimization with chemical accuracy.","authors":"Jeheon Woo, Seonghwan Kim, Jun Hyeong Kim, Woo Youn Kim","doi":"10.1038/s43588-025-00919-1","DOIUrl":"https://doi.org/10.1038/s43588-025-00919-1","url":null,"abstract":"<p><p>Here we introduce a framework for molecular structure optimization using a denoising model on a physics-informed Riemannian manifold (R-DM). Unlike conventional approaches operating in Euclidean space, our method leverages a Riemannian metric that better aligns with molecular energy change, enabling more robust modeling of potential energy surfaces. By incorporating internal coordinates reflective of energetic properties, R-DM achieves chemical accuracy with an energy error below 1 kcal mol<sup>-1</sup>. Comparative evaluations on QM9, QM7-X and GEOM datasets demonstrate improvements in both structural and energetic accuracy, surpassing conventional Euclidean-based denoising models. This approach highlights the potential of physics-informed coordinates for tackling complex molecular optimization problems, with implications for tasks in computational chemistry and materials science.</p>","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":" ","pages":""},"PeriodicalIF":18.3,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145892627","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-02DOI: 10.1038/s43588-025-00915-5
Shenghui Wu, Zhiwei Song, Xiang Zhang, Yifan Huang, Shuhang Chen, Xiang Shen, Jieyuan Tan, Mingdong Li, Ziyi Wang, Yujun Chen, Kai Liu, Dario Farina, Jose C Principe, Yiwen Wang
Prediction models that generate neuronal spikes from upstream neural activities offer a promising way to re-establish neural functional connectivity. Traditional methods train these models by supervised learning, which requires downstream recordings as ground truth. However, functional downstream activity cannot be recorded when neurological disorders exist. Here we introduce a reinforcement learning (RL)-based point process framework to generate spike trains that directly maximize behavior-level rewards, thus bypassing downstream recordings. This yields a generative spike model that directly transforms upstream activity into spike patterns modulated to desired behavior. We show that these RL-based generative models produce movement-modulated spike patterns akin to downstream recordings from healthy subjects, providing a biomimetic spike encoding framework. This RL framework outperforms existing methods and demonstrates a strong adaptation capability across different decoder settings, highlighting its potential for neural prostheses in restoring transregional communication with biomimetic cortical stimulation.
{"title":"A generative spike prediction model using behavioral reinforcement for re-establishing neural functional connectivity.","authors":"Shenghui Wu, Zhiwei Song, Xiang Zhang, Yifan Huang, Shuhang Chen, Xiang Shen, Jieyuan Tan, Mingdong Li, Ziyi Wang, Yujun Chen, Kai Liu, Dario Farina, Jose C Principe, Yiwen Wang","doi":"10.1038/s43588-025-00915-5","DOIUrl":"https://doi.org/10.1038/s43588-025-00915-5","url":null,"abstract":"<p><p>Prediction models that generate neuronal spikes from upstream neural activities offer a promising way to re-establish neural functional connectivity. Traditional methods train these models by supervised learning, which requires downstream recordings as ground truth. However, functional downstream activity cannot be recorded when neurological disorders exist. Here we introduce a reinforcement learning (RL)-based point process framework to generate spike trains that directly maximize behavior-level rewards, thus bypassing downstream recordings. This yields a generative spike model that directly transforms upstream activity into spike patterns modulated to desired behavior. We show that these RL-based generative models produce movement-modulated spike patterns akin to downstream recordings from healthy subjects, providing a biomimetic spike encoding framework. This RL framework outperforms existing methods and demonstrates a strong adaptation capability across different decoder settings, highlighting its potential for neural prostheses in restoring transregional communication with biomimetic cortical stimulation.</p>","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":" ","pages":""},"PeriodicalIF":18.3,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145893534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1038/s43588-025-00908-4
Tong Zhao (, ), Yan Zeng
Digital twins of self-driving chemistry laboratories may help reduce reliance on costly real-world experimentation and enable the testing of hypothetical automated workflows in silico.
{"title":"Digital twins for self-driving chemistry laboratories","authors":"Tong Zhao \u0000 (, ), Yan Zeng","doi":"10.1038/s43588-025-00908-4","DOIUrl":"10.1038/s43588-025-00908-4","url":null,"abstract":"Digital twins of self-driving chemistry laboratories may help reduce reliance on costly real-world experimentation and enable the testing of hypothetical automated workflows in silico.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 1","pages":"15-16"},"PeriodicalIF":18.3,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145879561","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1038/s43588-025-00924-4
Kourosh Darvish, Arjun Sohal, Abhijoy Mandal, Hatem Fakhruldeen, Nikola Radulov, Zhengxue Zhou, Satheeshkumar Veeramani, Joshua Choi, Sijie Han, Brayden Zhang, Jeeyeoun Chae, Alex Wright, Yijie Wang, Hossein Darvish, Yuchi Zhao, Gary Tom, Han Hao, Miroslav Bogdanovic, Gabriella Pizzuto, Andrew I. Cooper, Alán Aspuru Guzik, Florian Shkurti, Animesh Garg
Accelerated materials discovery is critical for addressing global challenges. However, developing new laboratory workflows relies heavily on real-world experimental trials, and this can hinder scalability because of the need for numerous physical make-and-test iterations. Here we present MATTERIX, a multiscale, graphics processing unit-accelerated robotic simulation framework designed to create high-fidelity digital twins of chemistry laboratories, thus accelerating workflow development. This multiscale digital twin simulates robotic physical manipulation, powder and liquid dynamics, device functionalities, heat transfer and basic chemical reaction kinetics. This is enabled by integrating realistic physics simulation and photorealistic rendering with a modular graphics processing unit-accelerated semantics engine, which models logical states and continuous behaviors to simulate chemistry workflows across different levels of abstraction. MATTERIX streamlines the creation of digital twin environments through open-source asset libraries and interfaces, while enabling flexible workflow design via hierarchical plan definition and a modular skill library that incorporates learning-based methods. Our approach demonstrates sim-to-real transfer in robotic chemistry setups, reducing reliance on costly real-world experiments and enabling the testing of hypothetical automated workflows in silico. MATTERIX, a multiscale graphics processing unit-accelerated framework for high-fidelity digital twins and workflows of chemistry laboratories, is presented, simulating robot and device operation, fluids and powders, and processes such as heat transfer and chemical kinetics.
{"title":"MATTERIX: toward a digital twin for robotics-assisted chemistry laboratory automation","authors":"Kourosh Darvish, Arjun Sohal, Abhijoy Mandal, Hatem Fakhruldeen, Nikola Radulov, Zhengxue Zhou, Satheeshkumar Veeramani, Joshua Choi, Sijie Han, Brayden Zhang, Jeeyeoun Chae, Alex Wright, Yijie Wang, Hossein Darvish, Yuchi Zhao, Gary Tom, Han Hao, Miroslav Bogdanovic, Gabriella Pizzuto, Andrew I. Cooper, Alán Aspuru Guzik, Florian Shkurti, Animesh Garg","doi":"10.1038/s43588-025-00924-4","DOIUrl":"10.1038/s43588-025-00924-4","url":null,"abstract":"Accelerated materials discovery is critical for addressing global challenges. However, developing new laboratory workflows relies heavily on real-world experimental trials, and this can hinder scalability because of the need for numerous physical make-and-test iterations. Here we present MATTERIX, a multiscale, graphics processing unit-accelerated robotic simulation framework designed to create high-fidelity digital twins of chemistry laboratories, thus accelerating workflow development. This multiscale digital twin simulates robotic physical manipulation, powder and liquid dynamics, device functionalities, heat transfer and basic chemical reaction kinetics. This is enabled by integrating realistic physics simulation and photorealistic rendering with a modular graphics processing unit-accelerated semantics engine, which models logical states and continuous behaviors to simulate chemistry workflows across different levels of abstraction. MATTERIX streamlines the creation of digital twin environments through open-source asset libraries and interfaces, while enabling flexible workflow design via hierarchical plan definition and a modular skill library that incorporates learning-based methods. Our approach demonstrates sim-to-real transfer in robotic chemistry setups, reducing reliance on costly real-world experiments and enabling the testing of hypothetical automated workflows in silico. MATTERIX, a multiscale graphics processing unit-accelerated framework for high-fidelity digital twins and workflows of chemistry laboratories, is presented, simulating robot and device operation, fluids and powders, and processes such as heat transfer and chemical kinetics.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 1","pages":"67-82"},"PeriodicalIF":18.3,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145879554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-22DOI: 10.1038/s43588-025-00895-6
Hengyun Zhou, Madelyn Cain, Mikhail D. Lukin
Quantum error correction provides a route to realizing large-scale quantum computation but incurs substantial resource overheads. Here we highlight recent advances that reduce these overheads by co-designing different levels of the computational stack, including algorithms, quantum-error-correction strategies and hardware architecture. We then discuss opportunities for further optimization such as leveraging flexible qubit connectivity and quantum low-density parity check codes. These strategies can bring useful quantum computation closer to reality as experiments advance in the coming years. Quantum error correction is vital for scalable quantum computing, but it incurs high resource overheads. This Perspective outlines recent breakthroughs and explores the opportunities to reduce the overheads by co-designing across algorithms, error-correction schemes and hardware architecture.
{"title":"Opportunities in full-stack design of low-overhead fault-tolerant quantum computation","authors":"Hengyun Zhou, Madelyn Cain, Mikhail D. Lukin","doi":"10.1038/s43588-025-00895-6","DOIUrl":"10.1038/s43588-025-00895-6","url":null,"abstract":"Quantum error correction provides a route to realizing large-scale quantum computation but incurs substantial resource overheads. Here we highlight recent advances that reduce these overheads by co-designing different levels of the computational stack, including algorithms, quantum-error-correction strategies and hardware architecture. We then discuss opportunities for further optimization such as leveraging flexible qubit connectivity and quantum low-density parity check codes. These strategies can bring useful quantum computation closer to reality as experiments advance in the coming years. Quantum error correction is vital for scalable quantum computing, but it incurs high resource overheads. This Perspective outlines recent breakthroughs and explores the opportunities to reduce the overheads by co-designing across algorithms, error-correction schemes and hardware architecture.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"5 12","pages":"1110-1119"},"PeriodicalIF":18.3,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802619","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-22DOI: 10.1038/s43588-025-00930-6
Bhupalee Kalita, Hatice Gokcan, Olexandr Isayev
As quantum mechanics marks its centennial in 2025, machine learning interatomic potentials have emerged as transformative tools in molecular modeling, bridging quantum mechanical accuracy with classical efficiency. Here we examine their development through four defining challenges—achieving chemical accuracy, maintaining computational efficiency, ensuring interpretability and reaching universal generalizability. We highlight architectural innovations, physics-informed approaches, and foundation models trained on extensive data. Together, these developments chart a path toward predictive, transferable and physically grounded machine learning frameworks for next-generation computational chemistry. As quantum mechanics marks its centennial, machine learning interatomic potentials are emerging as transformative tools bridging quantum accuracy with classical efficiency. This Perspective explores their evolution in terms of accuracy, efficiency, interpretability and generalizability challenges.
{"title":"Machine learning interatomic potentials at the centennial crossroads of quantum mechanics","authors":"Bhupalee Kalita, Hatice Gokcan, Olexandr Isayev","doi":"10.1038/s43588-025-00930-6","DOIUrl":"10.1038/s43588-025-00930-6","url":null,"abstract":"As quantum mechanics marks its centennial in 2025, machine learning interatomic potentials have emerged as transformative tools in molecular modeling, bridging quantum mechanical accuracy with classical efficiency. Here we examine their development through four defining challenges—achieving chemical accuracy, maintaining computational efficiency, ensuring interpretability and reaching universal generalizability. We highlight architectural innovations, physics-informed approaches, and foundation models trained on extensive data. Together, these developments chart a path toward predictive, transferable and physically grounded machine learning frameworks for next-generation computational chemistry. As quantum mechanics marks its centennial, machine learning interatomic potentials are emerging as transformative tools bridging quantum accuracy with classical efficiency. This Perspective explores their evolution in terms of accuracy, efficiency, interpretability and generalizability challenges.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"5 12","pages":"1120-1132"},"PeriodicalIF":18.3,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-22DOI: 10.1038/s43588-025-00942-2
As quantum mechanics marks its centennial, this issue of Nature Computational Science features a Focus that outlines the impact of quantum mechanics in advancing computing technologies, while discussing the challenges and opportunities that lie ahead.
{"title":"Reshaping computation with quantum mechanics","authors":"","doi":"10.1038/s43588-025-00942-2","DOIUrl":"10.1038/s43588-025-00942-2","url":null,"abstract":"As quantum mechanics marks its centennial, this issue of Nature Computational Science features a Focus that outlines the impact of quantum mechanics in advancing computing technologies, while discussing the challenges and opportunities that lie ahead.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"5 12","pages":"1093-1094"},"PeriodicalIF":18.3,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s43588-025-00942-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
First-principles electronic structure calculations have profoundly advanced research in physics, chemistry and materials science, yet their further development remains constrained by the accuracy–efficiency dilemma. Here we highlight recent breakthroughs in deep-learning methodologies that address this challenge, including the deep-learning quantum Monte Carlo method for the accurate study of correlated electrons and deep-learning density functional theory for efficient large-scale material simulations. These advances extend the reach of first-principles calculations to unprecedented scales and complexity, enhancing the impact of quantum mechanics in scientific discovery. This Review explores the integration of deep learning in first-principles electronic structure calculations, addressing the accuracy–efficiency dilemma of traditional algorithms and extending first-principles methods to unprecedented scales and complexity.
{"title":"Deep-learning electronic structure calculations","authors":"Zechen Tang, Haoxiang Chen, Yang Li, Yubing Qian, Yuxiang Wang, Weizhong Fu, Jialin Li, Chen Si, Wenhui Duan, Ji Chen, Yong Xu","doi":"10.1038/s43588-025-00932-4","DOIUrl":"10.1038/s43588-025-00932-4","url":null,"abstract":"First-principles electronic structure calculations have profoundly advanced research in physics, chemistry and materials science, yet their further development remains constrained by the accuracy–efficiency dilemma. Here we highlight recent breakthroughs in deep-learning methodologies that address this challenge, including the deep-learning quantum Monte Carlo method for the accurate study of correlated electrons and deep-learning density functional theory for efficient large-scale material simulations. These advances extend the reach of first-principles calculations to unprecedented scales and complexity, enhancing the impact of quantum mechanics in scientific discovery. This Review explores the integration of deep learning in first-principles electronic structure calculations, addressing the accuracy–efficiency dilemma of traditional algorithms and extending first-principles methods to unprecedented scales and complexity.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"5 12","pages":"1133-1146"},"PeriodicalIF":18.3,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号