Patterns, a Cell Press journal

Node persistence from topological data analysis reveals changes in brain functional connectivity This study employs persistent homology to investigate alterations in brain functional connectivity associated with healthy aging and autism spectrum disorder (ASD). Node persistence, a scalable local measure based on persistent homology introduced in this study, identifies brain regions linked to these conditions, including those with clinical evidence from non-invasive brain stimulation.

Online Now: Node persistence from topological data analysis reveals changes in brain functional connectivity #datascience

A self-supervised framework for emphysema anomaly detection and staging in computed tomography scans This study introduces a self-supervised learning framework that detects and stages emphysema from non-emphysema chest CT scans without manual labels. By modeling from non-emphysema lungs and using knowledge-guided synthetic patterns, the approach enables accurate, interpretable, and scalable assessment of emphysema in clinical imaging practice.

Online Now: A self-supervised framework for emphysema anomaly detection and staging in computed tomography scans #datascience

Spatial coherence in DNA barcode networks This study introduces spatial coherence, a way to check whether DNA barcode networks follow conventional geometry rules. Three geometry tests detect network quality, help with filtering corrections, and improve sequencing-based microscopy reconstructions in both simulations and published datasets.

Online Now: Spatial coherence in DNA barcode networks #datascience

Is AI overhyped? In this People of Data, we asked five researchers, including three members of the journal’s advisory board, whether they feel AI technologies are currently overhyped. Their responses reveal both optim...

Really enjoyed this Q&A @cp-patterns.bsky.social with five AI researchers, asking whether AI is overhyped. www.cell.com/patterns/ful...

Leveraging protein language models and a scoring function for indel characterization and transfer learning Insertions and deletions (indels) are difficult to study because they change protein sequence length, limiting existing tools. Here, IndeLLM is presented: a framework that leverages protein language models to assess indel effects in a more interpretable and generalizable way. The method maps indel impact to protein regions, improves predictive accuracy with a Siamese network, and provides guidelines for transfer learning. IndeLLM offers new opportunities for indel annotation and disease-related variant analysis.

Online Now: Leveraging protein language models and a scoring function for indel characterization and transfer learning #datascience

Confidence-weighted integration of human and machine judgments for superior decision-making When AI surpasses human performance, what can humans offer? We demonstrate that the performance of teams increases by integrating human judgments with those of machines. Integration is achieved by a straightforward regression approach that combines team members' confidence-weighted judgments.

Online Now: Confidence-weighted integration of human and machine judgments for superior decision-making #datascience

Predictive remapping and allocentric coding as consequences of energy efficiency in recurrent neural network models of active vision This study gives an example of how complex computations in neural networks can emerge from simple physical principles. Training a model to optimize internal energy efficiency while performing eye movements suffices for predictive remapping to emerge. The model learns to translate eye movements into an allocentric reference frame. Based on this reference frame, it learns to predict and inhibit the next fixation.

Online Now: Predictive remapping and allocentric coding as consequences of energy efficiency in recurrent neural network models of active vision #datascience

I've received some questions recently about the kinds of literature reviews we consider at @cp-patterns.bsky.social. Given also changes at the CS section of arXiv (blog.arxiv.org/2025/10/31/a...), I thought it would be timely to share some tips on submitting review papers to the journal. 1/n

Dignity, properly used, could be a useful construct in AI ethics Rueda et al. argue that the concept of dignity is problematic for AI ethics due to its complexity, ambiguity, and biased usage. While agreeing on many points, we propose that adding the necessary prec...

Great paper here from Cait Lamberton and colleagues replying to our previous publication in @cp-patterns.bsky.social

Dignity, properly used, could be a useful construct in AI ethics
www.cell.com/patterns/ful...

Is #AI overhyped? We asked five researchers, including three from @cp-patterns.bsky.social's Advisory Board. Here's what they thought:
www.cell.com/patterns/ful...

These two papers are examples of our commitment to covering research in non-English AI and natural language processing. For more, see our 2023 editorial
www.cell.com/patterns/ful...

Tucano: Advancing neural text generation for Portuguese Recent natural language processing (NLP) advances have favored high-resource languages while leaving many others underrepresented. This imbalance challenges global AI inclusivity. Addressing it requir...

Also in this issue, is a new resource for advancing natural language processing AI tools for Portuguese
www.cell.com/patterns/ful...

A network of nodes and edges encircles the globe, reflecting the worldwide reach of natural language processing as it spreads across languages and regions. From this global mesh, Greek letters emerge. When viewed horizontally, they are mirrored onto a continuously evolving background, suggesting the transformation of language technologies over time. The image highlights the work by Bakagianni et al. in this month’s issue, which offers a perspective on how machine learning for language has expanded dramatically across the world, affecting many languages, and captured here through the lens of Greek. Cover design by Smart Graphics, Ntina Gkanti. Image credit: Artistdesign.13 (shutterstock.com).

Our November issue is now live!
www.cell.com/patterns/iss...

On the cover this month is a colorful image highlighting the work by Bakagianni et al. that surveys natural language processing capabilities and tools for the Greek language
www.cell.com/patterns/ful...

New paper out! 🚨
“Assessing the adoption of the FAIR principles in Italian environmental research infrastructures”
We explore how 14 RIs are putting #FAIRdata into practice — key to building more open, reusable, and connected environmental science.
🧪 🌊 🧬 🪨

Contrastive learning enables epitope overlap predictions for targeted antibody discovery Antibodies are important for human health, but their diversity makes computational prediction of their binding properties challenging. Through contrastive fine-tuning of antibody language models on millions of antibody pairs, the authors enable accurate prediction of epitope overlap even among sequence-diverse antibodies, providing powerful new tools for therapeutic antibody discovery.

Online Now: Contrastive learning enables epitope overlap predictions for targeted antibody discovery #datascience

Assessing the adoption of the FAIR principles in Italian environmental research infrastructures The authors analyze how 14 environmental research infrastructures (RIs) implement the FAIR (Findable, Accessible, Interoperable, and Reusable) principles in Italy. The study uncovers diverse practices across the atmosphere, marine, biosphere, and geosphere subdomains. Results reveal not only strong heterogeneity but also emerging convergence, particularly in the marine subdomain. By highlighting transferable FAIR strategies, this work can guide other scientific communities interested in FAIR compliance and support the development of new digital ecosystems within and beyond Italy.

Online Now: Assessing the adoption of the FAIR principles in Italian environmental research infrastructures #datascience

Three-factor learning in spiking neural networks: An overview of methods and trends from a machine learning perspective This perspective surveys three-factor learning, a key mechanism for brain-inspired spiking neural networks. Adding a third, global signal, analogous to neuromodulators, allows these AI systems to learn and adapt more effectively. This overview connects recent advances in neuroscience and machine learning, outlining a path toward more powerful AI.

Online Now: Three-factor learning in spiking neural networks: An overview of methods and trends from a machine learning perspective #datascience

Sustainability challenges demand social science insights We live in an era of escalating sustainability challenges, such as exceeding the global 1.5°C threshold in 2024, the collapse of talks to address the widespread plastic crisis, and increasing wildfire...

In this month’s editorial, led by our social science editor Lu Liu, we highlight the importance of social sciences for sustainability. Social science has too often been seen as a soft science, a nice-to-have element or a useful add-on rather than a fundamental pillar

www.cell.com/cell-reports...

Are individuals who are positive about artificial intelligence also more unsure? As artificial intelligence matures, the impact it might have on how society functions is being actively pondered. In this opinion, through uniform-binomial mixtures, the author sheds quantitative ligh...

Are individuals who are positive about artificial intelligence also more unsure?: Applications of uniform-binomial mixed densities. Patterns www.cell.com/patterns/ful...

SIMBA: A robust and generalizable measure of data imbalance Data constitute machine learning’s fuel. Unfortunately, real-world data are often imbalanced. This poses a huge challenge to machine learning, as rare cases are hard to learn. So far, no adequate measure has been available to assess the impact of imbalance on machine learning’s performance. As a remedy for this omission, the authors introduce the status of imbalance (SIMBA), a reliable, robust, and generic imbalance measure that outperforms existing ones.

Online Now: SIMBA: A robust and generalizable measure of data imbalance #datascience

Decoding multi-joint hand movements from brain signals by learning a synergy-based neural manifold Brain-computer interfaces (BCIs) offer a promising avenue to restore motor function by converting neural activity into control signals for external devices. While current BCI technologies have successfully demonstrated brain-directed control of robotic arms, they still face significant limitations in dexterous hand movement control. This study proposes a novel approach called SynergyNet to decode complex hand movements directly from brain signals, facilitating high-performance BCI control with fine movements.

Online Now: Decoding multi-joint hand movements from brain signals by learning a synergy-based neural manifold #datascience

Text presented over a blue and orange background reads: Patterns, a Cell Press journal. Call for papers: Reanalysis. Compelling and creative reanalyses of prior works of high importance and broad impact. August 1, 2026.

New call for submissions!

To help promote critical reassessment of prior work, we are now inviting submissions that present compelling and creative reanalyses.

Learn more: www.cell.com/patterns/spe...

Submit before August 1, 2026

#openscience #datascience

Large language models for drug discovery and development In this review, the authors explore the transformative impact of large language models (LLMs) on drug discovery and development. They detail how LLMs can potentially accelerate our understanding of…

In this paper in @cp-patterns.bsky.social, Core Faculty member George Church, Postdoc Li Li, and their colleagues outline how LLMs are becoming crucial tools at every stage of drug development to dramatically reduce the time and cost associated with bringing new therapeutics to patients.

Computational workflows for natural and biomedical image processing based on hypercomplex algebras Quaternions, a type of hypercomplex number, are particularly useful in handling three-dimensional data, i.e., color images. Within the feature-rich hypercomplex setting, image-processing workflows can be realized for natural and biomedical images enabling alternative visual representations, offering effective solutions to current problems in computer vision and digital pathology, and generally expanding the scope and impact of hypercomplex image processing across a wide range of applications.

Online Now: Computational workflows for natural and biomedical image processing based on hypercomplex algebras #datascience

Unraveling learning characteristics of transformer models for molecular design Transformer models are adaptable to off-the-beaten-path molecular design tasks, such as protein-sequence-based compound design. The authors use these predictions as a model system to unravel the learning characteristics of transformers. The analysis reveals that the transformer predictions are statistically driven and that the models do not learn protein-ligand interactions or other biologically relevant information. Instead, compound memorization plays an important role. These findings caution against over-interpretation of sequence-based generative compound design using transformer models.

Online Now: Unraveling learning characteristics of transformer models for molecular design #datascience

In the image, proteins are represented as sailboats drifting in the dark. At the center, stands a lighthouse symbolizing the AI method LigUnity. Its beam illuminates several sailboats, guiding them toward glowing buoys, which symbolize ligands with high activity. The image illustrates the work by Feng et al. in this month's issue of Patterns. Image credit: Bin Feng, International Digital Economy Academy.

Our October issue is now live!
www.cell.com/patterns/iss...

Featured on the cover this month is a paper presenting LigUnity, a foundation model for predicting small molecule binding affinity for drug discovery and optimization applications www.cell.com/patterns/ful...

The widespread adoption of large language model-assisted writing across society This study explores AI-generated writing adoption across consumer complaints, corporate press releases, job postings, and international organization communications, revealing widespread use following ...

Very nice article co-authored by one of my former students, Mihai Codreanu, discovering how much of many documents eg press releases, job postings, have been written using LLMs (10-20% depending on type of document but it might have stabilised in late 2024) www.cell.com/patterns/ful...

The widespread adoption of large language model-assisted writing across society This study explores AI-generated writing adoption across consumer complaints, corporate press releases, job postings, and international organization communications, revealing widespread use following ChatGPT’s release. Findings highlight significant generative AI integration by small firms, corporations, and global institutions, emphasizing its transformative impact on communication practices, business strategies, and public policy.

Online Now: The widespread adoption of large language model-assisted writing across society #datascience

Neural mechanisms of visual quality perception and adaptability in the visual pathway Visual quality assessment is crucial for both human perception and artificial vision systems. This study reveals how the human brain naturally perceives image quality through distinct neural pathways. Functional magnetic resonance imaging results show that while low-level visual areas are highly sensitive to distortions, higher-level regions compensate to maintain recognition, much like how AI systems might “fill in” missing details. These findings not only deepen our understanding of human vision but also provide biological inspiration for developing more robust image-processing algorithms.

Online Now: Neural mechanisms of visual quality perception and adaptability in the visual pathway #datascience

Pan-microalgal dark proteome mapping via interpretable deep learning and synthetic chimeras The vast microbial “dark proteome” remains largely uncharacterized. Nelson et al. present LA4SR, a transformer framework that classifies algal sequences 10,000× faster than traditional methods with near-perfect recall. Trained on 77 million sequences including synthetic chimeras, the system can robustly classify unknown sequences with natural or scrambled start and stop information. They also developed an accompanying suite of explainability software to reveal biologically meaningful patterns linking amino acid patterns to evolutionary affiliations.

Online Now: Pan-microalgal dark proteome mapping via interpretable deep learning and synthetic chimeras #datascience