At this year’s SimTech Status Seminar, held to foster cross-disciplinary exchange within the SimTech community, Christian Pfaendner was awarded with the Best Paper Award. His paper, “ART-SM: Boosting Fragment-Based Backmapping by Machine Learning”, co-authored with Viktoria Korn (another SimTech doctoral researcher), Pritom Gogoi, Benjamin Unger, and Kristyna Pluhackova, was recognized for its methodological innovation, interdisciplinary collaboration, and reproducibility standards.
The award was presented during the seminar’s closing session, where Pfaendner gave an overview talk to the SimTech community and received a certificate and €500 in prize money, sponsored by the Industrial Consortium SimTech.
About the paper
Pfaendner’s work presents a novel machine learning–based framework for backmapping coarse-grained molecular dynamics simulations to atomistic resolution, significantly improving both accuracy and computational efficiency. The method – named ART-SM (Artificial Intelligence-based Reverse Transformation of Small Molecules) – builds on a fragment-based strategy and introduces a machine learning model to select the most suitable fragment conformations. This ensures that the reconstructed structures adhere closely to the Boltzmann distribution and represent realistic molecular behavior. The method has been validated on a variety of small molecules and a large micelle system. ART-SM is implemented in Python and the relevant code, tutorials, and data are publicly available on GitHub and DaRUS.
"Our approach bridges the gap between runtime-efficient coarse-grained simulations and detailed atomistic insights," said Pfaendner. "It enables seamless reconstruction of realistic molecular structures at the atomistic level - even for systems not explicitly seen during training, provided they are sufficiently similar to those encountered during training - with minimal user intervention."
In the future, the method will be extended to large biomacromolecules, materials, and drugs opening novel possibilities for high-resolution molecular structure reconstruction from low-resolution input.
Jury Statement
The jury, composed of SimTech Junior Professors and Independent Junior Research Group Leaders, highlighted three key reasons for selecting the paper. First, the work directly addresses a central SimTech research goals by combining simulation techniques with machine learning and bridging simulation scales. Second, it represents a successful interdisciplinary collaboration across multiple project networks—specifically PN3, PN4, and PN6—bringing expertise from chemistry, mathematics, machine learning and software development. Third, the paper sets a strong example in terms of scientific transparency: all source code, data, and documentation have been made publicly available via GitHub and DaRUS, ensuring full reproducibility.
In his laudatory speech, Junior Professor Marco Oesting stated that “this is a precise and technically solid piece of work. It clearly demonstrates the added value of interdisciplinary cooperation and addresses a relevant technical challenge in molecular simulations.”
Anneli Guthke, Independent Junior Research Group Leader and co-laudator, added: “The authors not only developed a novel approach but also made sure that it can be reused and understood by others – This is essential for long-term impact.”
“I’m glad that our method has been recognized. We focused on making it general, well-documented, and easy to use even for non-experts. That was important to me throughout the project“, underlines Christian Pfaendner after receiving the award.
SimTech Status Seminar and Support for Early-Career Researchers
The SimTech Status Seminar is held annually and brings together members of the Cluster from all project networks and disciplines. It provides a vital platform for doctoral researchers, postdocs, junior group leaders and professors to share results, build networks, and find new synergies.
SimTech supports early-career researchers through interdisciplinary supervision, technical training, and dedicated funding opportunities. The Best Paper Award highlights outstanding individual contributions and promotes transparency and collaboration within the cluster.
“The Status Seminar is more than just a get together – it brings together the SimTech community, highlights the community’s progress and functions as a platform for exchange and for future ideas,” said a representative from the SimTech executive board.