Benchmarking alignment methods for spatial transcriptomics data - Nature Computational Science A benchmark to evaluate spatial alignment methods for spatial transcriptomics shows that the best approach is dataset dependent and offers actionable guidance for the future development of methods.

📢New Analysis out today: a benchmark that systematically compares alignment methods for spatial transcriptomics and provides guidelines to help researchers select the optimal alignment method for their study. www.nature.com/articles/s43... 🖥️ 🧬

🔓 rdcu.be/fbBgo

Universal restoration of medical images - Nature Computational Science A self-supervised foundation model, HorusEye, learns realistic noise directly from X-ray scans and enables robust tomography restoration across diverse modalities, scanners, and tasks without clean tr...

📢Yide Zhang discusses the work of Xin Gao and colleagues on a self-supervised foundation model for restoring degraded X-ray tomographic images. www.nature.com/articles/s43... #Bioimaging

🔓 rdcu.be/fbBbT

SPRINGER NATURE: Open Science | Step-by-step guides. Share your research code and protocols openly. Best practices for transparency and reuse. Read the blogs. Abstract blue and purple design element on right.

Strengthening transparency and reproducibility starts with sharing research work openly.

Our latest blogs offer guidance for:
> How to share research code openly:
spklr.io/63329Eynzz
> How to share research protocols and methods openly:
spklr.io/63322EynzO

📢In the third opinion piece of our 5-year anniversary Series, Donald Truhlar and colleagues discuss some recent advances in density functional theory and propose future opportunities for the field. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/faqR5

A Möbius strip.

🚨Our March issue is now live, including an AI collaborator for science of science, a method for property-guided molecule generation, a Comment on the future of density functional theory, and much more! www.nature.com/natcomputsci...

📰Cover: www.nature.com/articles/s43... #cssky

Scaling and quantization of large-scale foundation model enables resource-efficient predictions in network biology - Nature Computational Science The authors demonstrate that the accuracy of predictions in network biology scales with larger foundation models pretrained with larger, more diverse data and that quantization enables resource-effici...

📢 Out now! Christina Theodoris and colleagues demonstrate that the accuracy of predictions in network biology scales with larger foundation models pretrained with larger, more diverse data, and that quantization enables resource-efficient predictions. www.nature.com/articles/s43... #NetSci 🧬🔄

HorusEye: a self-supervised foundation model for generalizable X-ray tomography restoration - Nature Computational Science HorusEye is a foundation model for universal X-ray tomography restoration that learns realistic degradation directly from data. It supports imaging at substantially lower doses and reduces hardware re...

📢 Xin Gao and colleagues develop HorusEye, a foundation model for universal X-ray tomography restoration, supporting low-dose image and reducing hardware requirements, while simultaneously improving quality. www.nature.com/articles/s43... #Bioimaging

🔓 rdcu.be/fapfh

De novo design of functional nucleic acids of aptamers - Nature Computational Science InstructNA leverages nucleic acid large language models with HT-SELEX for de novo generation of functional nucleic acids, exhibiting high efficiency and general applicability in designing aptamers for...

📢Da Han and colleagues present InstructNA, a framework that leverages nucleic acid LLMs and HT-SELEX to guide de novo design of functional nucleic acids without relying on structural information. www.nature.com/articles/s43...

BrainParc: unified lifespan brain parcellation from structural magnetic resonance images - Nature Computational Science This study introduces a unified framework for brain MRI tissue segmentation and region parcellation across the lifespan, demonstrating robust and consistent performance across heterogeneous datasets u...

📢Feng Shi, Dinggang Shen and colleagues introduce a framework for brain MRI tissue segmentation and region parcellation across the lifespan, showing consistent performance across heterogeneous datasets using a single model. www.nature.com/articles/s43... #compneurosky

🔓 rdcu.be/e7Txi

Choreographing molecular design with TANGO - Nature Computational Science A reward function (TANGO) is developed to enforce building blocks in generative artificial intelligence and leverage the synthesizability of high-value materials.

📢Tiago Rodrigues discusses the work by @pschwllr.bsky.social and Jeff Guo on TANGO, a reward function that augments molecular generative models to directly optimize for constrained synthesizability. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e7MRJ

Guiding molecular design with flow models - Nature Computational Science The PropMolFlow model uses flow matching to efficiently generate chemically valid molecules in three dimensions with targeted properties, enabling accelerated discovery of molecules useful in material...

📢Andreas Luttens discuss the work by @stemartiniani.bsky.social and colleagues on a flow-matching method for property-guided molecule generation. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e7BAy

A meaningful map of the underexplored electrolyte universe - Nature Computational Science A machine learning framework reveals how dynamic routing and interpretability can accelerate the discovery of better electrolytes for next-generation batteries.

📢Stephen Lam and Romakanta Bhattarai discuss the work by
Zhilong Wang and Fengqi You on a framework for modeling and interpreting salt-solvent chemistry. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e7Byy

High-throughput transition-state searches in zeolite nanopores - Nature Computational Science PoTS is an automated pipeline that maps reaction transition states inside zeolite pores. By identifying hundreds of confined transition states across many frameworks, it explains differences in cataly...

📢Out now! @rgblabmit.bsky.social presents PoTS, a pipeline for mapping reaction transition states inside zeolite pores, helping to explain the difference in catalytic selectivity and informing zeolite design. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e7Bsg

TANGO: direct optimization of constrained synthesizability for generative molecular design - Nature Computational Science The authors propose the TANGO reward function, which enables the generation of property-optimized small molecules with predicted synthesis routes, incorporating a small set of shared precursors.

📢💃It takes two to TANGO! Jeff Guo and @pschwllr.bsky.social present TANGO, a reward function that augments molecular generative models to directly optimize for constrained synthesizability. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e6C69

Understanding large-scale cooperation in a nonlinear, interconnected world - Nature Computational Science Large-scale cooperation is characterized by complex interaction patterns with nonlinear outcomes. Deepening our understanding may be critical to addressing real-world collective challenges.

📢In the second opinion piece of our 5-year anniversary Series, @evodynamics.bsky.social discusses the field of collective cooperation and the challenges ahead. www.nature.com/articles/s43... #evosky #cssky

🔓 rdcu.be/e5SIY

Cover runners-up of 2025 - Nature Computational Science We highlight some of our favorite cover suggestions from 2025.

📢We continue our yearly tradition and highlight our favorite author-suggested cover runners-up of 2025 in our latest Editorial! www.nature.com/articles/s43...

Flowing light with swirling patterns in motion.

🚨Our February issue is now live, including a generative spike-based framework to re-establish functional connectivity, a Comment on the future of large-scale cooperation, our cover runners-up of 2025, and much more! www.nature.com/natcomputsci...

Partially shared multi-modal embedding learns holistic representation of cell state - Nature Computational Science APOLLO is an autoencoder-based framework to integrate diverse data modalities while preserving both shared and modality-specific information. It enables predicting missing data modalities and identify...

📢Caroline Uhler and colleagues from Eric and Wendy Schmidt Center at @broadinstitute.org present APOLLO, a framework to integrate diverse data modalities, enabling predicting missing data modalities and identifying the influence of each modality on a phenotype. www.nature.com/articles/s43... 🖥️ 🧬

📢Rachel Adams and colleagues evaluate the potential and challenges for generative AI in LMICs, based on their own experience as part of a Gates Foundation program for developing projects that use LLM technologies in LMICs. www.nature.com/articles/s43... #ArtificialIntelligence

🔓 rdcu.be/e5kgy

Self-optimized spectral distance for low-light high-throughput Raman hyperspectral imaging - Nature Computational Science This work introduces an unsupervised method that restores high-quality Raman hyperspectral images from low-light measurements, enabling faster, lower-power imaging and expanding the use of Raman techn...

📢Hui Zhang and colleagues introduce a method for restoring high-quality Raman hyperspectral images from low-light measurements. www.nature.com/articles/s43... #ArtificialIntelligence #Bioimaging

🔓 rdcu.be/e5ka2

A neural network for modeling human concept formation, understanding and communication - Nature Computational Science The CATS Net framework models how abstract concepts emerge from sensory experience. Aligning with human brain activity and enabling knowledge transfer, it provides a unified framework for understandin...

📢Out now! Yang Chen, Yanchao Bi, Shan Yu and colleagues present a framework that models how abstract concepts emerge from sensory experience. www.nature.com/articles/s43... #compneurosky

A dynamic routing-guided interpretable framework for salt–solvent chemistry - Nature Computational Science This study presents a dynamic routing-guided framework to model and interpret salt–solvent chemistry, which effectively handles long-tailed data and captures the full spectrum of formulations, shaping...

📢Out now! Zhilong Wang and Fengqi You develop a framework called SCAN for modeling and interpreting salt-solvent chemistry. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e4JWQ

Learning the committor without collective variables - Nature Computational Science By learning directly from atomic motion, without the need for handcrafted descriptors, a graph neural network reveals how molecular systems change state, delivering accurate kinetics and atom-level in...

📢Out now! Work from @chipotlab.bsky.social and colleagues presents a method that learns from atomic motion without the need for handcrafted descriptors, allowing for accurate kinetics and atom-level insights. www.nature.com/articles/s43...

🔓 rdcu.be/e4p3B

Toward informed batch correction for single-cell transcriptome integration - Nature Computational Science Batch effects pose substantial challenges for obtaining meaningful biological insights from large-scale yet heterogeneous single-cell RNA-sequencing datasets. Here the authors review widely adopted ba...

📢New Perspective out! @penghe.bsky.social, @teichlab.bsky.social and colleagues review widely adopted batch correction methods and propose a path toward more informed, context-aware approaches for future method development. www.nature.com/articles/s43... 🖥️ 🧬

🔓 rdcu.be/e4pUJ

Deep learning for asymmetric catalysis - Nature Computational Science A recent study develops a model for predicting stereoselectivity and absolute configurations in asymmetric hydrogenation of olefins.

📢News & Views out today! Robert Paton and @sun9823.bsky.social discuss the paper by Bo Zhang and colleagues on a deep learning approach that can predict the sense and magnitude of enantioselectivity in asymmetric hydrogenations. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e3jFC

DiffSyn: a generative diffusion approach to materials synthesis planning - Nature Computational Science A generative AI approach is developed for predicting materials synthesis recipes—a complex challenge in materials science. Using this approach, the authors experimentally synthesized a material using ...

📢New Article alert! Elsa Olivetti and colleagues develop DiffSyn, a generative approach for predicting materials synthesis recipes. www.nature.com/articles/s43... #chemsky

🔓 rdcu.be/e1ViT

Improving atlas-scale single-cell annotation models with hierarchical cross-entropy loss - Nature Computational Science A hierarchical cross-entropy loss is presented, which incorporates ontology structure into training and improves the out-of-distribution performance of large-scale single-cell annotation models withou...

📢Out now! Sebastiano Cultrera di Montesano, Peter S. Winter, @lcrawford.bsky.social, and colleagues present a hierarchical cross-entropy loss that improves performance of single-cell annotation models. www.nature.com/articles/s43... 🖥️ 🧬

deepmriprep: voxel-based morphometry preprocessing via deep neural networks - Nature Computational Science deepmriprep leverages neural networks to enable voxel-based morphometry preprocessing of MRI data that is 37× faster than existing methods while achieving comparable accuracy in segmentation, registra...

📢Out now! @codingfisch.bsky.social and colleagues present deepmriprep, a tool that leverages neural networks to enable 37x faster Voxel-based Morphometry preprocessing of MRI data than existing methods. www.nature.com/articles/s43... #compneurosky

Many more commentary pieces are coming this year for our special Series! Stay tuned!

🧵(4/4)

The evolution of digital twins from reactive to agentic systems - Nature Computational Science Digital twins are evolving into self-learning, autonomous systems that link models, data and human interaction. Realizing their full potential depends on interoperability, standardization and the inte...

In the first opinion piece of the Series, Omer San and colleagues discuss the timely topic of digital twins and their evolution from analytical instruments to autonomous systems. www.nature.com/articles/s43...

🧵(3/4)