Nature computational science最新文献_第9页

HELIX: a scalable model for predicting context-dependent regulation of RNA splicing and isoform usage HELIX：预测RNA剪接和同工异构体使用的上下文依赖调节的可扩展模型。

IF 18.3

Nature computational science Pub Date : 2026-05-19 DOI: 10.1038/s43588-026-00988-w

Zihan Zhou, Bingqi Wu, Xin Zheng, Lipu Song, Shuai Zhang, Dali Han, Zhaoqi Liu, Yuan Gao

{"title":"HELIX: a scalable model for predicting context-dependent regulation of RNA splicing and isoform usage","authors":"Zihan Zhou, Bingqi Wu, Xin Zheng, Lipu Song, Shuai Zhang, Dali Han, Zhaoqi Liu, Yuan Gao","doi":"10.1038/s43588-026-00988-w","DOIUrl":"10.1038/s43588-026-00988-w","url":null,"abstract":"Context-dependent alternative splicing plays a critical role in disease pathogenesis and organ development, but its complex regulation remains challenging to predict. Here, to address this, we developed HELIX, a hierarchical deep learning framework that integrates pre-mRNA sequence and RNA-binding protein expression profiles to predict tissue- and condition-specific splicing patterns and transcript isoform usage simultaneously. By leveraging both short-read and long-read RNA sequencing data during training, HELIX achieves greater accuracy than existing splicing prediction models and conventional short-read–based methods in predicting differential splicing events, splicing strength at highly regulated splice sites, and isoform usage. The model enables systematic identification of tissue-specific splicing quantitative trait loci and their functional impacts. Furthermore, HELIX predicts patient-specific splicing dysregulation with quantitative attribution to genetic variants and abnormal RNA-binding protein expression in colon cancer cohorts. Through transfer learning, the HELIX model can be adapted to single-cell RNA sequencing data, thereby enabling the prediction of cell-type-specific isoforms. HELIX is a deep learning framework for modeling tissue- and context-specific RNA splicing and isoform usage, prioritizing splicing-altering genetic variants and their functional effects from bulk and single-cell RNA sequencing data.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 5","pages":"464-477"},"PeriodicalIF":18.3,"publicationDate":"2026-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147977774","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}

引用次数: 0

Modeling context-dependent RNA splicing by deep learning 通过深度学习建模上下文依赖的RNA剪接。

IF 18.3

Nature computational science Pub Date : 2026-05-19 DOI: 10.1038/s43588-026-00987-x

Chengxuan Chen (, ), Leng Han (, )

引用次数: 0

HESpotEx: a dual-stream deep learning framework for spot-level gene expression prediction from histological images HESpotEx：一个双流深度学习框架，用于从组织学图像中预测点水平的基因表达。

IF 20.3

Nature computational science Pub Date : 2026-05-15 DOI: 10.1038/s43588-026-00992-0

Wang Yin, Qin Peng, Fanyi Meng, You Wan, Weilong Zhang, Yuan Zhou

{"title":"HESpotEx: a dual-stream deep learning framework for spot-level gene expression prediction from histological images","authors":"Wang Yin, Qin Peng, Fanyi Meng, You Wan, Weilong Zhang, Yuan Zhou","doi":"10.1038/s43588-026-00992-0","DOIUrl":"10.1038/s43588-026-00992-0","url":null,"abstract":"Whole-slide histopathological images (WSIs) constitute a fundamental approach in disease diagnosis and prognosis. Recently emerging spatial transcriptomics (ST) methods can reveal the spatial gene expression landscape behind the histopathological images, but with much higher cost. Here, therefore, we propose HESpotEx, a dual-stream multimodal deep learning framework to predict the spatial gene expression patterns solely from WSI images. Leveraging graph attention autoencoders, an image encoder and a graph convolution network decoder, HESpotEx is capable of predicting expressions of up to 5,457 genes across individual spatial sampling spots from WSIs. HESpotEx exhibits superior performance and better robustness on ST datasets from various cancer and noncancer samples as well as on a large-scale The Cancer Genome Atlas WSI dataset. Moreover, on our in-house WSI dataset, HESpotEx also underscores diagnosis-associated WSI patches. Finally, HESpotEx shows better cross-sectional consistency in the latest high-resolution ST datasets. Together, our results demonstrate the potential of HESpotEx to decipher the spatial molecular characteristics underlying tissue histological patterns. This work reports HESpotEx, an artificial intelligence model that predicts spot-level spatial gene expression from whole-slide histopathology images to capture the tissue molecular patterns across cancers and inflammatory diseases.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 8","pages":"842-855"},"PeriodicalIF":20.3,"publicationDate":"2026-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147950492","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}

引用次数: 0

Scaling and quantization of a foundational deep learning model for network biology 网络生物学基础深度学习模型的缩放和量化。

IF 18.3

Nature computational science Pub Date : 2026-05-11 DOI: 10.1038/s43588-026-00990-2

引用次数: 0

SmileyLlama: modifying large language models for directed chemical space exploration 为定向化学空间探索修改大型语言模型。

IF 20.3

Nature computational science Pub Date : 2026-05-11 DOI: 10.1038/s43588-026-00986-y

Joseph M. Cavanagh, Kunyang Sun, Andrew Gritsevskiy, Dorian Bagni, Yingze Wang, Thomas D. Bannister, Teresa Head-Gordon

{"title":"SmileyLlama: modifying large language models for directed chemical space exploration","authors":"Joseph M. Cavanagh, Kunyang Sun, Andrew Gritsevskiy, Dorian Bagni, Yingze Wang, Thomas D. Bannister, Teresa Head-Gordon","doi":"10.1038/s43588-026-00986-y","DOIUrl":"10.1038/s43588-026-00986-y","url":null,"abstract":"Here we show that large language models (LLMs) can be transformed via supervised fine-tuning of engineered prompts into SmileyLlama for exploring the chemical space of drug molecules. We benchmark SmileyLlama against pretrained LLMs and chemical language models trained from scratch for generating valid and novel drug-like molecules, and use direct preference optimization to both improve SmileyLlama’s adherence to a prompt and as part of the iMiner reinforcement learning framework to predict molecules with optimized three-dimensional conformations and high binding affinity to drug targets. By training an LLM to speak directly as a chemical language model, while retaining most of its natural language capabilities, we show that SmileyLlama can reliably generate molecules with user-specified properties rather than acting only as a chatbot with knowledge of chemistry or as a virtual assistant. While SmileyLlama is geared toward drug discovery, the supervised fine-tuning/direct preference optimization/LLM framework can be extended to other chemical, biological and materials applications. A general-purpose large language model can be trained to create a specialized model, SmileyLlama, that can generate drug-like molecules with user-specified properties that are relevant to medicinal chemistry through a process that might be extended to other scientific domains.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 8","pages":"856-869"},"PeriodicalIF":20.3,"publicationDate":"2026-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s43588-026-00986-y.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147876734","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}

引用次数: 0

Translating biodiversity into chemical diversity 将生物多样性转化为化学多样性。

IF 20.3

Nature computational science Pub Date : 2026-05-05 DOI: 10.1038/s43588-026-00984-0

引用次数: 0

DeepSeMS: revealing the hidden biosynthetic potential of the global ocean microbiome with a large language model deepsem：通过大型语言模型揭示全球海洋微生物群隐藏的生物合成潜力。

IF 20.3

Nature computational science Pub Date : 2026-04-30 DOI: 10.1038/s43588-026-00983-1

Tingjun Xu, Yuwei Yang, Ruixin Zhu, Weili Lin, Jixuan Li, Yan Zheng, Peng Zhang, Guoqing Zhang, Guoping Zhao, Na Jiao

{"title":"DeepSeMS: revealing the hidden biosynthetic potential of the global ocean microbiome with a large language model","authors":"Tingjun Xu, Yuwei Yang, Ruixin Zhu, Weili Lin, Jixuan Li, Yan Zheng, Peng Zhang, Guoqing Zhang, Guoping Zhao, Na Jiao","doi":"10.1038/s43588-026-00983-1","DOIUrl":"10.1038/s43588-026-00983-1","url":null,"abstract":"Microbial-derived secondary metabolites (SMs) hold great therapeutic potential but are predominantly discovered from cultured species, representing only a fraction of microbial biodiversity. Advances in metagenomics have unveiled reservoirs of biosynthetic gene clusters (BGCs), but translating genomic sequences into precise chemical structures remains challenging owing to the structural complexity of cryptic BGCs and the context-dependent substrate tolerance and cross-reactivity of modular biosynthetic domains. Here we present DeepSeMS, a transformer-based large language model that accurately predicts secondary metabolite chemical structures from BGC sequences. By encoding biosynthetic genes as functional domains and leveraging a feature-aligned data augmentation, DeepSeMS outperformed existing methods and successfully generated chemically valid predictions for 96.38% of cryptic BGCs. Applying DeepSeMS to a global ocean metagenome, we characterized over 60,000 secondary metabolites, revealing chemical diversity, ecological specificity and considerable biomedical potential, especially as antibiotics. This study underscores the capability of deep learning-driven approaches in revealing hidden biosynthetic potential of Earth’s largest, yet largely unexplored, microbial ecosystem. This work presents DeepSeMS, a transformer-based framework that translates biosynthetic gene clusters into candidate secondary metabolite structures, enabling scalable exploration of chemical diversity from uncultivated global ocean microbiomes.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 7","pages":"753-766"},"PeriodicalIF":20.3,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147823877","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}

引用次数: 0

What reviewers request the most 评论者最要求的是什么

IF 18.3

Nature computational science Pub Date : 2026-04-28 DOI: 10.1038/s43588-026-00989-9

引用次数: 0

NOEM: efficient and scalable finite element method enabled by reusable neural operators NOEM：由可重复使用的神经算子实现的高效和可扩展的有限元方法

IF 18.3

Nature computational science Pub Date : 2026-04-28 DOI: 10.1038/s43588-026-00974-2

Weihang Ouyang, Yeonjong Shin, Si-Wei Liu, Lu Lu

{"title":"NOEM: efficient and scalable finite element method enabled by reusable neural operators","authors":"Weihang Ouyang, Yeonjong Shin, Si-Wei Liu, Lu Lu","doi":"10.1038/s43588-026-00974-2","DOIUrl":"10.1038/s43588-026-00974-2","url":null,"abstract":"The finite element method (FEM) is a well-established numerical method for solving partial differential equations (PDEs). However, its mesh-based nature gives rise to substantial computational costs, especially for complex multiscale simulations. Emerging machine learning-based methods provide data-driven solutions to PDEs, yet they present challenges, including high training cost and low model reusability. Here we propose the neural-operator element method (NOEM) by synergistically combining FEM with operator learning to address these challenges. NOEM leverages neural operators to simulate subdomains that require fine meshes in FEM. In each subdomain, a neural operator is used to build a single element, namely, a neural-operator element (NOE). NOEs are then integrated with standard finite elements to represent the entire solution through the variational framework. Thereby, NOEM does not necessitate dense meshing and offers efficient simulations. We demonstrate the accuracy, efficiency and scalability of NOEM by performing systematic theoretical analysis and numerical experiments, such as nonlinear PDEs, multiscale problems, PDEs on complex geometries and discontinuous coefficient fields. The authors propose a method that unifies finite element methods and machine learning by using neural operators as elements to model complex subdomains, yielding an efficient and scalable numerical framework with highly reusable machine learning models.","PeriodicalId":74246,"journal":{"name":"Nature computational science","volume":"6 4","pages":"417-429"},"PeriodicalIF":18.3,"publicationDate":"2026-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147754495","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}

引用次数: 0

Merging the classical and the modern in physics-based simulations 在基于物理的模拟中融合经典和现代

IF 18.3

Nature computational science Pub Date : 2026-04-28 DOI: 10.1038/s43588-026-00978-y

Michael D. Shields, Somdatta Goswami

引用次数: 0