Shiota, Ishihara, and Mizukami demonstrate that intermediate information in a universal neural network potential can be used to create a general-purpose descriptor for chemical property prediction, in the latest featured #DigitalDiscovery article. Find out more in the open access paper: https://lnkd.in/eQtqVKgB
Digital Discovery #RSCDigital
Non-profit Organizations
Machine learning, robotics, and AI - discover the latest developments in digital molecular discovery from the RSC
About us
Machine learning, robotics, and AI - discover the latest developments in digital molecular discovery from the RSC.
- Website
-
https://www.rsc.org/journals-books-databases/about-journals/
External link for Digital Discovery #RSCDigital
- Industry
- Non-profit Organizations
- Company size
- 501-1,000 employees
Updates
-
In our latest featured article, Singla, Zhao, and Savoie show how kinetic data of alkanes under pyrolysis can be used to teach a model that predicts a scalar measure of the thermal stability from the molecular graph alone. Read the open access paper in #DigitalDiscovery: https://lnkd.in/dWBz94cB
-
Digital Discovery #RSCDigital reposted this
Massive congratulations to Erika Magnusson (Chalmers University of Technology) and Benjamin Xu Shi (University of Cambridge), the poster prize winners at the Correlated electronic structure Faraday Discussion held last week in London ✨ Erika's poster was titled: "Multireference error mitigation to tackle strongly correlated electronic structure problems on quantum hardware" Benjamin's poster was titled: "It’s not a S(KZ)CAM! Many-body methods for surface chemistry have come of age!" PCCP and the Royal Society of Chemistry were delighted to sponsor this prices
-
Our next featured #DigitalDiscovery article, from Cho, Corminboeuf et al., provides a method to rapidly assign ground state spin to TM complexes in cell2mol, for the efficient exploitation of crytallographic datasets. Read the open access paper here: https://lnkd.in/ebpYSPnY
-
Digital Discovery #RSCDigital reposted this
Shine a light on your energy materials research. The Joint Chemical Science RSC-CSJ symposium 2024 is your chance to show us what you’re working on and connect with the wider Chemical Science, Energy #RSCEnergy and Materials #RSCMat communities – submit your poster abstract by 19 August #ChemSciSymposium https://lnkd.in/gCiaG_Bd
-
Digital Discovery #RSCDigital reposted this
🌟 Join us at the Computational Molecular Science (CMS 2024) Conference! 🌟 📅 Dates: 11-13 September 📍 Location: University of Warwick Be part of the 7th edition of the biennial CMS series, bringing together leading computational experts in bio and soft matter modelling, ab initio calculations, and machine learning over two and a half days. Keynote Speakers: Karsten Reuter -- Fritz-Haber-Institut der Max-Planck-Gesellschaft Noa Marom -- Carnegie Mellon University Marjolein Dijkstra -- Utrecht University Sarah Anne Harris -- University of Leeds Michele Ceriotti -- EPFL Invited Speakers: Tom Penfold, Ralf Tonner-Zech, Grant Hill, Rachel Crespo Otero, Flor Siperstein, David Wilkins, Carla de Tomas, Daniel Cole, Antonia Mey, Clotilde Cucinotta, and Joe Greener Conference Highlights: -Evening poster session with buffet dinner -Early-career networking event -Closing with the Royal Society of Chemistry Theoretical Chemistry Group's Graduate student meeting 🔔 Register soon to secure on-campus accommodation for the two nights! We still have a few slots for contributed talks and are accepting poster abstracts. Don't miss this opportunity to connect and collaborate with leading researchers in the field. Register Now! https://lnkd.in/gVT6SD4V #CMS2024 #ComputationalScience #MolecularScience #UniversityofWarwick #Networking Sponsored by Royal Society of Chemistry, CP5, CCP9, CCPBioSim Biomolecular Simulation, Institute of Advanced Study, University of Warwick, EPSRC HetSys-CDT programme Co-organised by Livia Bartok-Partay, Svenja M. Janke and myself -- University of Warwick - Department of Chemistry
-
EuroQSAR 2024 is coming soon! Abstract submission for posters closes on 31 July 2024. The RSC's journals Digital Discovery, Molecular Systems Design & Engineering, Analyst, PCCP, and RSC Medicinal Chemistry are proud to be sponsoring awards for the best poster this year - don't miss your opportunity to take part! Find out more: https://www.euroqsar.org/
-
Our next featured #DigitalDiscover article is a Perspective, exploring informatics-based methods for the design of nuclear waste forms. Saal et al. demonstrate the advantages of a system design approach leveraging a variety of approaches, and use it in the design of ceramic waste forms. Read the open access paper here: https://lnkd.in/eq4zTRdi
-
Bing Yan shares our latest featured #DigitalDiscovery article: "We introduce CatScore, a learning-based evaluation metric tailored for asymmetric catalysis in organic chemistry. CatScore enables highly efficient and effective evaluation of diverse catalyst design models at both the instance and system levels". Read the open access paper here: https://lnkd.in/eq4zTRdi
-
Digital Discovery #RSCDigital reposted this
I am glad to announce that our Multimodal co-orchestration article has been published in #DigitalDiscovery. Here we introduce a framework to expedite the investigation and characterization of combinatorial libraries and similar systems by co-orchestration of multiple modalities (tools). The method integrates dimensionality reduction through variational autoencoders (VAEs) with multi-task Gaussian Processes (GPs) to harness correlations between modalities, thereby accelerating exploration. Leveraging VAEs enables us to transfer the raw data interpretation from humans to AI agents, allowing GPs to be trained on the compositional dependencies of complex features without requiring direct human interpretation. My endless gratitude to co-authors Utkarsh Pratiush, Ilia Ivanov, Yongtao Liu, Rohit Pant, Xiaohang Zhang, Ichiro Takeuchi, Maxim Ziatdinov and Sergei Kalinin.
In our next featured #DigitalDiscovery article, Slautin, Kalinin et al. demonstrate the benefits of a co-orchestration approach to automated instrumentation, "leveraging knowledge about a compositional correlation uncovered for one property to expedite the exploration of another property measured by a different method, thereby accelerating the overall characterization process". Their paper shows this highly general approach applied to piezoresponse force microscopy and micro-Raman spectroscopy on a combinatorial Sm-BiFeO3 library. Read the open access article here: https://lnkd.in/es48WjpX