Watch the video of Kevin Jablonka's talk: "Why machine learning can find a new material, but not a needle in a haystack". Delivered in the FAIRmat Seminar on September 15, 2023.

Clara P. Marshall, Julia Schumann and Annette Trunschke published the viewpoint article “Achieving Digital Catalysis: Strategies for Data Acquisition, Storage and Use” in Angewandte Chemie.

Considering the scientific principles of catalysis, this article discusses the requirements for the future research data infrastructure. A paradigm shift in data acquisition and management is needed to address the challenges in developing high-performance, stable, scalable, and cost-effective catalysts for energy storage and sustainable, climate-neutral chemical synthesis.

The paper “MOFGalaxyNet: a social network analysis for predicting guest accessibility in metal–organic frameworks utilizing graph convolutional networks” by our Mehrdad Jalali, Dinga Wonanke and Christof Wöll was published open access in the Journal of Cheminformatics.

This study presents the newly designed tool MOFGalaxyNet, a galaxy-like social network. In combination with a Graphical Convolutional Network (GCN), MOFGalaxyNet predicts the guest accessibility of a given MOF, which is a key performance parameter for this material class. MOFGalaxyNet provides a robust approach for screening MOFs for host-guest interaction studies.

This CECAM flagship workshop, co-organized by the FAIRmat consortium and the MoSDeF group, hosted by the Max Planck Institute for Polymer Research, brought together a diverse group of people within the molecular dynamics (MD) community: members of FAIR-data (or similar) consortia, developers of MD simulation engines, university or institutional data stewards, HPC resource managers, and researchers in the field particularly interested in the development of FAIR-data standards.

As one of the first workshops of its kind within this community, this event was designed to be discussion-focused, to make concrete proposals for addressing the significant challenges in the “Fair-ification” of molecular simulations. In addition to the invited talks and associated Q&A sessions, 5 distinct round-table discussions were held, including a final wrap-up discussion to brainstorm how to continue discussion and collaboration within the community. The discussion topics also included data provenance strategies for MD simulations, storage of simulation (meta)data, interoperability of simulation engines, and data structures for edge cases.

FAIRmat was not only represented in terms of the internal organizers and representatives from our computational area but also by our new Area C PI, Prof. Dr. Tristan Bereau, who showcased his group’s dataset of C₇O₂ isomers, which was used to parametrize a transferable coarse-grained force field for binary mixtures. This dataset represents the most extensive set of molecular dynamics simulations currently held in the NOMAD repository and demonstrates the potential of NOMAD as a helpful tool for the MD community.

The meeting ended with an optimistic perspective for future gatherings and collaborations between existing software projects. FAIRmat looks forward to participating in the continued developments of FAIR-data management within this community!

Our NOMAD software is now published in The Journal of Open Source Software, an open-access journal for research software packages. NOMAD is a web-based application that provides data management for materials science research data. 

From September 12-14, 2023, the Association Nationale Forschungsdateninfrastuktur (NFDI) hosted the 1st conference on research data infrastructure (CoRDI) at the Karlsruhe Institute of Technology. These three days brought together all of the NFDI consortia and research data management (RDM) experts from around the world to exchange ideas and connect our community. 

Team members representing several different FAIRmat Areas attended. FAIRmat contributed to the conference with two talks and two posters. Heiko Weber (leader of Area B: Experiment) presented a talk titled “Research Data Management for Experiments in Solid-State Physics: Concept”, and Markus Scheidgen (Infrastructure coordinator) presented a talk titled “FAIR research data with NOMAD”. Two FAIRmat poster contributions were also presented: “FAIRmat guide to writing data management plans” by Ahmed Mansour (coordinator of Area F: User support, training, and outreach) and  “Towards FAIR data in heterogeneous catalysis research” by Julia Schumann (catalysis expert in Area E: Use case demonstrators).

In addition to the oral and poster presentations, our outreach team was also present at the FAIRmat stand at the “Marketplace of the consortia”, where we chatted with the conference participants about FAIRmat and NOMAD and explored possible opportunities for collaboration.

To view our contributions, you can visit the FAIRmat community on Zenodo. To view other contributions to the conference, check out the conference proceedings or the CoRDI community on Zenodo. 

The 2nd CoRDI will take place in 2025, we look forward to meeting you there!

Our sibling consortia NFDI4Chem awards the FAIRest dataset in chemistry! The FAIR4Chem Award honors researchers in chemistry who publish their research datasets according to the FAIR principles, thus making a significant contribution to increasing transparency in research and the reuse of scientific knowledge. Apply until November 17, 2023.

Watch the video of Maia G. Vergniory's talk, originally delivered in the FAIRmat seminar on July 28, 2023.

On July 25, 2023, we launched the very first FAIRmat podcast series!

In this series, called Pioneers in electronic-structure theory, the FAIRmat team talks with personalities who were instrumental in the development of methods and codes which have come to be known as electronic-structure theory. This includes pioneering work in density-functional theory and methods beyond to treat many-body effects in ground-state and excitations. For our first episode, Miguel Marques talks with Hardy Groß about the early days of density-functional theory, the development of time-dependent DFT (TDDFT)together with Erich Runge, his days in Santa Barbara with Walter Kohn, and much more.

Find the podcast on YouTube or enjoy the audio-only version on Spotify now, and subscribe to make sure the second episode (coming soon!) appears in your feed.

Taylor D. Sparks' talk "Materials informatics: Moving beyond screening via generative machine learning models", given in the Physical Sciences in NFDI colloquium on June 15, 2023, is now available to watch on our YouTube channel!