RSC - Digital Discovery latest articles

Benchmarking explainable AI methods for toxicophore detection and toxicity prediction

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00576K, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Dina Khasanova, Igor V. Tetko
High-accuracy models yield stable and chemically meaningful XAI explanations, with multiple methods consistently identifying functional groups and toxicophores across prediction tasks.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Constraint-aware labware layout generation from natural language for heterogeneous laboratory robots

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00026F, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Yuya Tahara-Arai, Takashi Inagaki, Akari Kato, Koji Ochiai, Kazuya Azumi, Koichi Takahashi, Genki N. Kanda, Haruka Ozaki
The Labware-Layout Planner takes a natural language experimental description, an available labware list, and constraints as input to generate a comprehensive labware placement map and a set of natural language instructions.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Identification of multi-transcriptomic prognostic biomarkers to explore natural therapeutics for lung cancer integrating machine learning

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00045B, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Md Ahad Ali, Hriddhi Sarker, Marguba Kamrun, Humaira Sheikh, Bilkis Akter Shifa, Siam Ahmed, Tarikul Islam, Sujoy Banik, Neeraj Kumar
Transcriptomics and machine learning approaches spotlight natural compounds as CDK1 blockers.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Meta-learning linear models for molecular property prediction

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00443H, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Yulia Pimonova, Michael G. Taylor, Alice Allen, Ping Yang, Nicholas Lubbers
A meta-learning linear approach exploits small, separate datasets to improve property predictions in the low-data regime.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Molecular Arms Race Classifier for Decrypting Venom Peptide and Ion Channel Interactions

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D5DD00498E, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Favour T Achimba, Arezoo Bybordi, Mariam Gelashvili, Jessy Ramirez, Anita Raja, Weigang Qiu, Mande Holford
Animal venoms comprise an astonishing number of peptides, proteins and small molecules. The diversity of venom compounds arises from evolutionary adaptations, resulting in both offensive and defensive traits in predators...
The content of this RSS Feed (c) The Royal Society of Chemistry

Assessing the Extrapolation Capability of Template-free Retrosynthesis Models

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D6DD00072J, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Yousung Jung, Jonghwi Choe, Shuan Chen
Template-free retrosynthesis models offer the potential to extrapolate beyond established chemical reaction spaces, addressing inherent limitations of template-based approaches. However, it remains unclear whether these models can reliably predict accurate,...
The content of this RSS Feed (c) The Royal Society of Chemistry

Identification of drug candidates against glioblastoma with machine learning and high-throughput screening of heterogeneous cellular models

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D5DD00190K, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Vanessa Smer-Barreto, Richard Elliott, John C. Dawson, Álvaro Lorente-Macías, Muhammad Furqan, Asier Unciti-Broceta, Diego Oyarzún, Neil Carragher
Glioblastoma multiforme (GBM) is an aggressive primary brain tumour that presents significant treatment challenges due to its complex pathology and heterogeneity. The lack of validated molecular targets is a major...
The content of this RSS Feed (c) The Royal Society of Chemistry

CREOLab: A Procedure Captioning Dataset for Understanding Creative Tool Use in Object-Rich Laboratory Videos

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D5DD00542F, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Shigeaki Goto, Tatsuki Hasebe
“Creative tool use” refers to the flexible application of tools beyond their intended purpose. In scientific experiments, this behavior is described as a “lab hack,” and its automatic documentation is...
The content of this RSS Feed (c) The Royal Society of Chemistry

Uncertainty-aware active learning reveals reliability limits in lead-free halide perovskite screening

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00153J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Xiyao Yu
Uncertainty-aware active learning maps reliability limits in lead-free halide perovskite screening, guiding high-fidelity calculations.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Accelerating discovery across scientific disciplines through reproducible workflows with AiiDAlab

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00567A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Aliaksandr V. Yakutovich, Daniel Hollas, Edan Bainglass, Jusong Yu, Corsin Battaglia, Miki Bonacci, Lucas Fernandez Vilanova, Stephan Henne, Anders Kaestner, Michel Kenzelmann, Graham Kimbell, Jakob Lass, Fabio Lopes, Daniel G. Mazzone, Andres Ortega-Guerrero, Xing Wang, Nicola Marzari, Carlo A. Pignedoli, Giovanni Pizzi
AiiDAlab provides a Jupyter-based environment for reproducible computational workflows powered by AiiDA provenance tracking, integrating ELNs and experimental facilities to enable data-driven scientific discovery across a broad range of disciplines.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Stoichiometrically-informed symbolic regression for extracting chemical reaction mechanisms from data

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00470E, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Manuel Palma Banos, Joel D. Kress, Rigoberto Hernandez, Galen T. Craven
A data-driven computational method is introduced to extract chemical reaction mechanisms from time series chemical concentration data.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

AI-designed and AI-implemented Control Systems for Bespoke Scientific Instrumentation: Application to Scanning Microscopy

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D6DD00177G, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Ke Wang, David James Ward, Matthew Ord, Boyao Liu, Andrew Peter Jardine
The pace of innovation in custom scientific instrumentation is frequently bottlenecked by the complexity of software engineering. While hardware designs evolve rapidly, developing robust and integrated control systems remains resource-intensive...
The content of this RSS Feed (c) The Royal Society of Chemistry

Vision-guided adaptive scooping for powder weighing in autonomous chemistry laboratories

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00569H, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Nikola Radulov, Thomas Little, Andrew I. Cooper, Gabriella Pizzuto
The integration of vision-guided perception into the powder scooping stage allowed our system to autonomously correct acquisition errors, enabling robust weighing of heterogeneous materials for solid-state materials chemistry workflows.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Reaction Center Prediction by Analyzing Attention of a Chemical Language Model

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D6DD00055J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Xiaoliang Xiong, Ruizhen Jia, Yang Tian, Jingke Chen, Boxue Tian
Pretrained chemical language models are widely used to predict molecular properties and chemical reactions, yet interpreting their internal attention mechanisms remains difficult. Here, we analyze attention matrices from a chemical...
The content of this RSS Feed (c) The Royal Society of Chemistry

How Digital Is Chemical Research? Insights from the Second NFDI4Chem Community Survey on Research Data and FAIR Workflows

Digital Discovery, 2026, Accepted Manuscript
DOI: 10.1039/D5DD00584A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Jochen Ortmeyer, Vitali Sidorin, Daniela Adele Hausen, Ann-Christin Andres, John D Jolliffe, Theo Bender, Giacomo Lanza, Steffen Neumann, Oliver Koepler, Johannes C. Liermann, Sonja Herres-Pawlis, Nicole Jung, Christoph Steinbeck
Increasing digitisation is revolutionising all scientific disciplines, but it poses a particular challenge in chemistry. Research data is generated, for example, during the synthesis of substances, the recording of spectra...
The content of this RSS Feed (c) The Royal Society of Chemistry

Benchmarking physics-inspired machine learning models for transition metal complexes with diverse charge and spin states

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00571J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Yuri Cho, Ksenia R. Briling, Yannick Calvino Alonso, Ruben Laplaza, Clemence Corminboeuf
We benchmark two classes of physics-inspired machine learning models for predicting quantum-chemical properties of transition metal complexes using three complementary datasets.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Contributors to the Digital Discovery Emerging Investigators collection 2025

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD90017H, Profile

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

This article profiles the early career researchers whose work features in the Digital Discovery Emerging Investigators collection 2025.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

First annual Digital Discovery Emerging Investigators collection

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD90016J, Editorial

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

This collection showcases research carried out by internationally recognised, up-and-coming scientists in the early stage of their independent careers who are making outstanding contributions to their respective fields.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

CReM-dock: de novo design of synthetically feasible structures guided by molecular docking

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00131A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Guzel Minibaeva, Haolin Du, Finlay Clark, Julien Michel, Pavel Polishchuk
The de novo generation of chemical compounds represents a compelling strategy for the exploration of a significantly broader chemical space compared to traditional virtual screening methods.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

AiiDA-TrainsPot: towards automated training of neural-network interatomic potentials

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00005C, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Davide Bidoggia, Nataliia Manko, Maria Peressi, Antimo Marrazzo
AiiDA-TrainsPot is an automated, modular active-learning workflow that efficiently trains neural-network interatomic potentials with minimal human supervision.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

ChatMat: a multi-agent chemist for autonomous material prediction and exploration

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00582E, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Shuai Lv, Lei Peng, Shizhe Jiao, Yufan Yao, Wentiao Wu, Wei Hu
ChatMat is an autonomous multiagent framework integrating LLMs with DFT and ML-PES to accelerate material property prediction and exploration. By coordinating specialized agents, it executes complex workflows with minimal human intervention.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Discovery of hydrogen storage molecules using large language models and machine learning

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00102E, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Hassan Harb, Magali S. Ferrandon, Timothy A. Goetjen, Seryeong Lee, Omar K. Farha, Massimiliano Delferro, Rajeev Surendran Assary
Accelerating the discovery of new molecules with targeted properties is a central challenge in molecular design.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Structured domain knowledge enables trustworthy materials science question-answering with large language models

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00028B, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Daegun Lee, Jiwoo Choi, Gyeong Hoon Yi, Seok Su Sohn, Byungju Lee, Donghun Kim
The domain-aligned question-answering system based on structured database construction and RAG pipeline.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

RAISE: a self-driving laboratory for interfacial property formulation discovery

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00531K, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Mohammad Nazeri, Sheldon Mei, Jeffrey Watchorn, Alex Zhang, Erin Ng, Tao Wen, Abhijoy Mandal, Kevin Golovin, Alán Aspuru-Guzik, Frank Gu
RAISE is the first autonomous closed-loop system linking formulations to surface wettability to discover formulations from user-defined goals.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Autonomous sampling and SHAP interpretation of deposition-rates in bipolar HiPIMS

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00063K, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Alexander Wieczorek, Nathan Rodkey, Jan Sommerhäuser, Jason Hattrick-Simpers, Sebastian Siol
Autonomous loops and SHAP analysis allow for comprehensive, statistical analysis of >3000 different bipolar HiPIMS processes.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Python-controlled, solvent-resistant fraction collector for automated flow synthesis

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00554J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Hongchen Wang, Owen A. Meville, Harrison A. Mills, Monique Ngan, Jay R. Werber, Nipun Kumar Gupta
We present a low-cost, solvent-resistant, open-source fraction collector that delivers real-time, volume-based sampling for automated flow chemistry.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

WeChemSynOntology: semantic modeling of wet chemical syntheses in a self-driving lab for nano- and advanced materials

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00058D, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Markus Schilling, Harald Bresch, Bernd Bayerlein, Bastian Ruehle
The WCSO formally describes wet chemical syntheses, enabling precise, machine-readable representations that improve reproducibility, sharing, querying, and autonomous workflow execution across self-driving labs and materials acceleration platforms.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Enhancing predictive modeling with molecular fingerprint fusion strategies

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00302D, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Viktoriia Turkina, Melanie R. W. Messih, Etienne Kant, Jelle T. Gringhuis, Annemieke Petrignani, Garry Corthals, Jake W. O'Brien, Saer Samanipour
QSAR performance depends on molecular representation; mid-level fusion of non-hashed fingerprints may improve accuracy across diverse tasks.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

FlowMol3: flow matching for 3D de novo small-molecule generation

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00363F, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Ian Dunn, David R. Koes
A generative model capable of sampling realistic molecules with desired properties could accelerate chemical discovery across a wide range of applications.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

The loss landscape of powder X-ray diffraction-based structure optimization is too rough for gradient descent

Digital Discovery, 2026, 5,1590-1599
DOI: 10.1039/D6DD00017G, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Nofit Segal, Akshay Subramanian, Mingda Li, Benjamin Kurt Miller, Rafael Gómez-Bombarelli
While potential energy surfaces offer smooth convergence, we show that XRD similarity metrics create a highly non-convex, ill-posed loss landscape. This rugged topology severely complicates gradient-based crystal structure optimization.
The content of this RSS Feed (c) The Royal Society of Chemistry

Multi-stage Bayesian optimisation for dynamic decision-making in self-driving labs

Digital Discovery, 2026, 5,1900-1912
DOI: 10.1039/D5DD00572H, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Luca Torresi, Pascal Friederich
A structure-aware, resumable framework enables dynamic pausing of multi-stage experiments via intermediate measurements. MSBO identifies optimal solutions more efficiently than traditional Bayesian optimisation via data-driven resource allocation.
The content of this RSS Feed (c) The Royal Society of Chemistry

On-the-fly fine-tuning of foundational neural network potentials: a Bayesian neural network approach

Digital Discovery, 2026, 5,1845-1867
DOI: 10.1039/D5DD00392J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Tim Rensmeyer, Denis Kramer, Oliver Niggemann
We introduce a workflow for data-efficient on-the-fly finetuning of foundational neural network potentials with enhanced trustworthiness by harnessing predictive uncertainty estimation extracted from a Bayesian transfer learning approach.
The content of this RSS Feed (c) The Royal Society of Chemistry

Large language models for porous materials: from text mining to autonomous laboratory

Digital Discovery, 2026, 5,1470-1500
DOI: 10.1039/D5DD00578G, Review Article

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Seunghee Han, Taeun Bae, Junho Kim, Younghun Kim, Jihan Kim
Schematic overview of LLM integration in porous materials research, including NLP-based text mining, LLM adaptation, and autonomous laboratory systems.
The content of this RSS Feed (c) The Royal Society of Chemistry

Integrating machine learning interatomic potentials with batched optimization for crystal structure prediction

Digital Discovery, 2026, 5,1913-1924
DOI: 10.1039/D6DD00016A, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Chengxi Zhao, Zhaojia Ma, Dingrui Fan, Siyu Hu, Leping Wang, Feng Hua, Weile Jia, En Shao, Guangming Tan, Jun Jiang, Linjiang Chen
BOMLIP-CSP is an open-source framework that accelerates crystal structure prediction via batched optimization with MLIPs while maintaining accuracy.
The content of this RSS Feed (c) The Royal Society of Chemistry

SALSA: a low-cost self-driving lab modular add-on for salt solubility assessment for battery electrolytes

Digital Discovery, 2026, 5,1881-1887
DOI: 10.1039/D5DD00516G, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Tianyi Zhang, Hongyi Lin, Yuhan Chen, Venkatasubramanian Viswanathan
We propose a low-cost automated salt solubility screening module. Using an excess-solvent workflow with a 3D-printed Archimedes-screw solid doser and computer-vision dissolution detection, it enables faster mapping of electrolyte solubility trends.
The content of this RSS Feed (c) The Royal Society of Chemistry

NaviDiv: a web app for monitoring chemical diversity in generative molecular design

Digital Discovery, 2026, 5,1579-1589
DOI: 10.1039/D5DD00487J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Mohammed Azzouzi, Thanapat Worakul, Clémence Corminboeuf
The rapid progress in generative models for molecular design has led to extensive libraries of candidate molecules for biological and chemical applications.
The content of this RSS Feed (c) The Royal Society of Chemistry

Evaluation of foundational machine learned interatomic potentials for migration barrier predictions

Digital Discovery, 2026, 5,1809-1819
DOI: 10.1039/D5DD00534E, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Achinthya Krishna Bheemaguli, Penghao Xiao, Gopalakrishnan Sai Gautam
We benchmark foundational MLIPs against DFT-NEB calculated migration barriers and the corresponding accuracy of intermediate geometries. We also highlight the possible correlation between the accuracy of barrier and geometry predictions.
The content of this RSS Feed (c) The Royal Society of Chemistry

Fast and scalable retrosynthetic planning with a transformer neural network and speculative beam search

Digital Discovery, 2026, 5,1783-1793
DOI: 10.1039/D5DD00573F, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Natalia Andronova, Mikhail Andronov, Jürgen Schmidhuber, Michael Wand, Djork-Arné Clevert
Medusa model and speculative beam search accelerate template-free synthesis planning.
The content of this RSS Feed (c) The Royal Society of Chemistry

Learning potential energy surfaces of hydrogen atom transfer reactions in peptides

Digital Discovery, 2026, 5,1831-1844
DOI: 10.1039/D6DD00054A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Marlen Neubert, Patrick Reiser, Frauke Gräter, Pascal Friederich
We build a DFT dataset for peptide hydrogen atom transfer and train atomistic ML potentials to learn reactive potential energy surfaces, enabling accurate reaction barrier prediction from energy evaluations.
The content of this RSS Feed (c) The Royal Society of Chemistry

POLARIS: perovskite optimization using LLM-assisted refinement and intelligent screening

Digital Discovery, 2026, 5,1765-1782
DOI: 10.1039/D5DD00378D, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Jordan Marshall, Sheryl L. Sanchez, Rushik Desai, Elham Foadian, Utkarsh Pratiush, Arun Mannodi-Kanakkithodi, Sergei V. Kalinin, Mahshid Ahmadi
LLM-assisted mining converts 200 studies on 2D halide perovskites into a spacer-cation dataset. GNN-GP models predict structural and optoelectronic targets with uncertainty, and clustering reveals chemical spacer families.
The content of this RSS Feed (c) The Royal Society of Chemistry

A symmetry-preserving and transferable representation for learning the Kohn–Sham density matrix

Digital Discovery, 2026, 5,1868-1880
DOI: 10.1039/D5DD00230C, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Liwei Zhang, Patrizia Mazzeo, Michele Nottoli, Edoardo Cignoni, Lorenzo Cupellini, Benjamin Stamm
We present a parameterized representation for learning the mapping from a molecular configuration to its corresponding density matrix using the atomic cluster expansion (ACE) framework, which preserves the physical symmetries of the mapping.
The content of this RSS Feed (c) The Royal Society of Chemistry

High-performance training and inference for deep equivariant interatomic potentials

Digital Discovery, 2026, 5,1558-1567
DOI: 10.1039/D5DD00423C, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Chuin Wei Tan, Marc L. Descoteaux, Mit Kotak, Gabriel de Miranda Nascimento, Seán R. Kavanagh, Laura Zichi, Menghang Wang, Aadit Saluja, Yizhong R. Hu, Tess Smidt, Anders Johansson, William C. Witt, Boris Kozinsky, Albert Musaelian
The NequIP framework is redesigned for scalable distributed training and PyTorch 2.0 compilation. AOT Inductor inference and optimized Allegro kernels accelerate molecular dynamics by factors of 5–18 on practical system sizes.
The content of this RSS Feed (c) The Royal Society of Chemistry

Hacking 3D printers as laboratory robots

Digital Discovery, 2026, 5,1460-1469
DOI: 10.1039/D5DD00451A, Review Article

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Sander Baas, Nessa Carson, Vittorio Saggiomo
The emergence of affordable 3D printers has enabled laboratories to optimize setups, print custom parts, and accelerate research. A new movement has emerged over the past decade, in which 3D printers are repurposed as laboratory-specific robots.
The content of this RSS Feed (c) The Royal Society of Chemistry

ComProScanner: a multi-agent based framework for composition-property structured data extraction from scientific literature

Digital Discovery, 2026, 5,1794-1808
DOI: 10.1039/D5DD00521C, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Aritra Roy, Enrico Grisan, John Buckeridge, Chiara Gattinoni
ComProScanner is an end-to-end multi-agentic autonomous platform that extracts, validates and visualises machine-readable compositions, properties and synthesis data from journal articles via Text and Data Mining APIs for database creation.
The content of this RSS Feed (c) The Royal Society of Chemistry

Increasing trustworthiness of machine learning-based drug sensitivity prediction with a multivariate random forest approach

Digital Discovery, 2026, 5,1746-1764
DOI: 10.1039/D5DD00284B, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Lisa-Marie Rolli, Lea Eckhart, Lutz Herrmann, Andrea Volkamer, Hans-Peter Lenhof, Kerstin Lenhof
MORGOTH is a comprehensive approach that unites simultaneous regression and classification, interpretability, reliability, and robustness in a single framework. We demonstrate its capabilities to predict anti-cancer drug responses.
The content of this RSS Feed (c) The Royal Society of Chemistry

Deep graph kernel learning for material & atomic level uncertainty quantification in adsorption energy prediction

Digital Discovery, 2026, 5,1568-1578
DOI: 10.1039/D6DD00020G, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Osman Mamun, Chenlu Yang, Shuwen Yue
Deep graph kernel learning combines GNN backbones with Gaussian processes to deliver calibrated, atomic-level uncertainty quantification for adsorption energy prediction, outperforming ensemble and evidential methods for catalyst discovery.
The content of this RSS Feed (c) The Royal Society of Chemistry

ALBATROSS: a robotised system for high-throughput electrolyte screening via automated electrolyte formulation, coin-cell fabrication, and electrochemical evaluation

Digital Discovery, 2026, 5,1522-1530
DOI: 10.1039/D5DD00580A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Hyun-Gi Lee, Jaekyeong Han, Minjun Kwon, Hyeonuk Kwon, Jooha Park, Hoe Jin Hah, Dong-Hwa Seo
ALBATROSS is an automated platform for high-throughput electrolyte screening, integrating electrolyte formulation, coin-cell assembly, and electrochemical evaluation.
The content of this RSS Feed (c) The Royal Society of Chemistry

Enhancing high-dimensional neural network potential accuracy in OLED systems via element relabeling

Digital Discovery, 2026, 5,1820-1830
DOI: 10.1039/D5DD00579E, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Yonghwan Yun, Dongmin Park, Junyoung Choi, Dong Shin Choi, Yousung Jung
Accurate atomistic simulations are essential for understanding organic light-emitting diode (OLED) materials in complex condensed-phase environments.
The content of this RSS Feed (c) The Royal Society of Chemistry

Reaction optimization through mechanistic insight and predictive modelling

Digital Discovery, 2026, 5,1447-1459
DOI: 10.1039/D5DD00543D, Review Article

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Roger Monreal-Corona, Anna Pla-Quintana, Albert Poater
The optimization of chemical reactions lies at the heart of synthetic chemistry. We review how DoE, computational models, and machine learning drive progress in efficiency and selectivity.
The content of this RSS Feed (c) The Royal Society of Chemistry

Assessment of molecular dynamics time series descriptors in protein–ligand affinity prediction

Digital Discovery, 2026, 5,1888-1899
DOI: 10.1039/D5DD00452G, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Jakub Poziemski, Artur Yurkevych, Pawel Siedlecki
Generation of time-based descriptors from protein–ligand molecular dynamics simulations improves generalization and reveals limits of MD-driven affinity prediction.
The content of this RSS Feed (c) The Royal Society of Chemistry

InSpecLearn4SDL: interpretable spectral features predict conductivity in self-driving doped conjugated polymer labs

Digital Discovery, 2026, 5,1925-1947
DOI: 10.1039/D5DD00479A, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Ankush Kumar Mishra, Jacob P. Mauthe, Nicholas Luke, Aram Amassian, Baskar Ganapathysubramanian
To accelerate materials discovery using self-driving labs (SDLs), we present a machine learning pipeline that predicts the electrical conductivity of doped conjugated polymers using rapid, non-destructive optical spectroscopy.
The content of this RSS Feed (c) The Royal Society of Chemistry

Machine learning inversion of interatomic force constants from single-crystal inelastic neutron scattering

Digital Discovery, 2026, 5,1545-1557
DOI: 10.1039/D6DD00008H, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Aiden Sable, Bander Linjawi, Kyle Bradbury, Jordan Malof, Olivier Delaire
Machine learning bridges synthetic and experimental inelastic neutron scattering data to recover interatomic force constants under realistic measurement conditions.
The content of this RSS Feed (c) The Royal Society of Chemistry

Database utility for cyclovoltammetry knowledge (DUCK): unified platform for electrochemical data

Digital Discovery, 2026, 5,1736-1745
DOI: 10.1039/D6DD00019C, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Diego Garay-Ruiz, Sergio Pablo-García, Han Hao, Marisol Martín-González
The DUCK platform enables the FAIR management and visualization of cyclic voltammetry data via knowledge graphs.
The content of this RSS Feed (c) The Royal Society of Chemistry

Looking back and to the future after four-plus years of language in chemistry

Digital Discovery, 2026, 5,1440-1446
DOI: 10.1039/D5DD00550G, Perspective

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Glen M. Hocky, Andrew D. White
We discuss the promise of large language models in chemistry four years ago and the strides, expected and unexpected, that have been made in the intervening years. This image generated by Google Gemini.
The content of this RSS Feed (c) The Royal Society of Chemistry

Exploring the deviation from Nernst–Einstein conductivity in ionic liquids using machine learning

Digital Discovery, 2026, 5,1709-1717
DOI: 10.1039/D5DD00414D, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Aditi Seshadri, Lyndon T. M. Hess, Shuwen Yue
Correcting the Nernst–Einstein equation using sigma profile-based machine learning models is an accurate, interpretable approach to estimate ionic liquid conductivities.
The content of this RSS Feed (c) The Royal Society of Chemistry

Machine learning models for catalytic asymmetric reactions of simple alkenes: from enantioselectivity predictions to chemical insights

Digital Discovery, 2026, 5,1718-1735
DOI: 10.1039/D5DD00483G, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Ajnabiul Hoque, Nupur Jain, Divya Chenna, Raghavan B. Sunoj
Enantioselectivity predictions of catalytic asymmetric transformations of alkenes using a class imbalance-aware AttentiveFP machine learning model offer good accuracy and chemically meaningful interpretability.
The content of this RSS Feed (c) The Royal Society of Chemistry

Pessimistic asynchronous sampling in high-cost Bayesian optimization

Digital Discovery, 2026, 5,1613-1622
DOI: 10.1039/D5DD00477B, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Amanda A. Volk, Kristofer G. Reyes, Jeffrey G. Ethier, Luke A. Baldwin
Pessimistic model predictions in asynchronous Bayesian optimization can enable more efficient and robust experimental system optimization in both asychronous and serial sampling settings.
The content of this RSS Feed (c) The Royal Society of Chemistry

Jeweler-in-the-loop: personalized alloy color optimization via preference-based BO

Digital Discovery, 2026, 5,1700-1708
DOI: 10.1039/D6DD00009F, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Chase Katz, Ting-Yu Yang, Parker King, Md Shafiqul Islam, Brent Vela, Raymundo Arróyave
We introduce a jeweler-in-the-loop framework combining Thermo-Calc optical modeling with preference-based Bayesian optimization to personalize alloy color design.
The content of this RSS Feed (c) The Royal Society of Chemistry

Distilling system complexity to enable unbiased and predictive computational reaction investigations

Digital Discovery, 2026, 5,1531-1544
DOI: 10.1039/D5DD00366K, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Raphaël Robidas, Claude Y. Legault
Automated and predictive computational investigations are possible with systems up to 42 reactive atoms through novel strategies of pruning reaction possibilities.
The content of this RSS Feed (c) The Royal Society of Chemistry

Coupled fragment-based generative modeling with stochastic interpolants

Digital Discovery, 2026, 5,1600-1612
DOI: 10.1039/D5DD00535C, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Tuan Le, Yanfei Guan, Djork-Arné Clevert, Kristof T. Schütt
Fragment-based generative model using stochastic interpolants for structure-based drug design. Explicit conditional training outperforms inpainting (87% vs. 25%). PLK3 kinase validation confirms diverse fragment generation with key interactions.
The content of this RSS Feed (c) The Royal Society of Chemistry

A universal machine learning model for the electronic density of states

Digital Discovery, 2026, 5,1635-1649
DOI: 10.1039/D5DD00557D, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Wei Bin How, Pol Febrer, Sanggyu Chong, Arslan Mazitov, Filippo Bigi, Matthias Kellner, Sergey Pozdnyakov, Michele Ceriotti
Leveraging the expressive PET architecture and chemically diverse MAD dataset, PET-MAD-DOS enables rapid and accurate evaluation of the electronic density of states and bandgap across diverse chemical spaces, from molecules to bulk crystals.
The content of this RSS Feed (c) The Royal Society of Chemistry

Using Flory–Huggins-informed human-in-the-loop Bayesian optimization to map the phase diagram of polymer blends

Digital Discovery, 2026, 5,1675-1688
DOI: 10.1039/D5DD00556F, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Justin C. Hughes, Dylan J. York, Kevin G. Yager, Chinedum O. Osuji, Russell J. Composto
Integrating Flory–Huggins theory into a Bayesian optimization workflow enhances the objectivity of mapping phase behavior in polymer blends.
The content of this RSS Feed (c) The Royal Society of Chemistry

Siamese graph neural networks for melting temperature prediction of molten salt eutectics

Digital Discovery, 2026, 5,1689-1699
DOI: 10.1039/D5DD00480B, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Nila Mandal, James Maniscalco, Mark Aindow, Qian Yang
A Siamese GNN predicts eutectic melting temperatures from molecular structure without engineered features, improves with data augmentation from single-component training data, and enables transfer from inorganic molten salts to organic eutectics.
The content of this RSS Feed (c) The Royal Society of Chemistry

Structure-guided machine learning for efficiency prediction of organic photovoltaics using experimentally informed molecular descriptors

Digital Discovery, 2026, 5,1510-1521
DOI: 10.1039/D5DD00496A, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

JuHyun Lee, HyoJin Ban, HyunIl Seo, HangKen Lee, Fiza Arshad, DongWook Kim
Machine learning prediction combining structural descriptors and experimental information accurately models ternary organic solar cells (D1 : D2 : A). An XGBoost model achieves an R² of 0.849 for data-driven discovery of organic photovoltaics.
The content of this RSS Feed (c) The Royal Society of Chemistry

Synthesis planning in reaction space: a study on success, robustness and diversity

Digital Discovery, 2026, 5,1623-1634
DOI: 10.1039/D5DD00280J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Alan Kai Hassen, Helen Lai, Samuel Genheden, Mike Preuss, Djork-Arné Clevert
Reaction-centric synthesis planning reduces search space redundancy. Online search outperforms self-play and is more robust to stock changes. Route diversity saturates across algorithms, making the single-step model, not search, the key limitation.
The content of this RSS Feed (c) The Royal Society of Chemistry

Development of accurate transferable hydrofluorocarbon refrigerant force fields using a machine learning and optimization approach

Digital Discovery, 2026, 5,1650-1674
DOI: 10.1039/D5DD00537J, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Montana N. Carlozo, Ning Wang, Alexander W. Dowling, Edward J. Maginn
Transferable hydrofluorocarbon force fields that accurately predict thermophysical property data can be rapidly postulated and parameterized with Gaussian process models.
The content of this RSS Feed (c) The Royal Society of Chemistry

Statistics makes a difference: machine learning adsorption dynamics of functionalized cyclooctyne on Si(001) at DFT accuracy

Digital Discovery, 2026, 5,1501-1509
DOI: 10.1039/D5DD00420A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Hendrik Weiske, Rhyan Barrett, Ralf Tonner-Zech, Patrick Melix, Julia Westermayr
The interpretation of experiments on reactive semiconductor surfaces requires statistically significant sampling of molecular dynamics, but conventional ab initio methods are limited due to prohibitive computational costs.
The content of this RSS Feed (c) The Royal Society of Chemistry

Chemist Eye: a visual language model-powered system for safety monitoring and robot decision-making in self-driving laboratories

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00062B, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Francisco Munguia-Galeano, Zhengxue Zhou, Satheeshkumar Veeramani, Hatem Fakhruldeen, Louis Longley, Rob Clowes, Andrew I. Cooper
Safety risks in self-driving laboratories (SDLs) are amplified by robotics and automation, motivating the development of real-time AI-driven monitoring and hazard-response systems.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Generalization of long-range machine learning potentials in complex chemical spaces

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00570A, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Michał Sanocki, Julija Zavadlav
Long-range interactions are key to making machine learning interatomic potentials accurate and transferable across chemical space.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Large language models in materials science and the need for open-source approaches

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00499C, Review Article

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Fengxu Yang, Weitong Chen, Jack D. Evans
This work explores the use of large language models across data mining, chemical understanding, predictive insight, and autonomous experimental orchestration.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

Masgent: an AI-assisted materials simulation agent

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D6DD00043F, Paper

Open Access

This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.

Guangchen Liu, Songge Yang, Yu Zhong
Masgent is an AI-driven materials simulation agent that translates user intent into automated workflows. Integrating DFT, MLPs, and ML with a feedback loop, it streamlines simulations and accelerates data-driven materials discovery.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

A deep learning approach to searching property spaces of materials

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00422E, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Robert J. Appleton, Brian C. Barnes, Steven F. Son, Alejandro Strachan
Large scale discovery of melt-castable molecular materials using deep learning property prediction and genetic algorithms.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry

MolRes-DTA: a molecular-multiview fusion and residue-aware model for drug-target affinity prediction

Digital Discovery, 2026, Advance Article
DOI: 10.1039/D5DD00365B, Paper

Open Access

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.

Hongli Hou, Qi Wei, Dian Huang, Minglu Zhao, Hongliang Duan, Shengzhong Feng
MolRes-DTA fuses drug and protein features for affinity prediction.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry