View All Molecular Dynamics at Instruct
The Molecular Bioinformatics & Computational Biophysical Chemistry facility at Forschungszentrum Jülich (IBG-4) provides advanced expertise in structural and molecular bioinformatics, integrating computational biophysics with data-driven approaches to study biomolecular systems.
The service focuses on modeling, simulation, and AI-based analysis to relate biomolecular structure, dynamics, interactions, and function, supporting rational design of proteins and enzymes. As part of Instruct-ERIC, it offers tailored computational support that complements experimental structural biology through integrative, structure-guided research strategies. Rather than standardized workflows, the service involves close collaboration, individualized method development, and hands-on scientific guidance for selected projects.
Access to the facility is generally available during standard business hours, with opportunities for on-site collaboration as well as coordinated remote work following initial engagement.
The facility has experience of using the supercomputers JURECA, JUWELS, and JUPITER at Jülich Supercomputing Centre, which are available for projects on a computing grant basis.
The facility operates a versatile, high-performance compute cluster designed to support both AI-driven and physics-based molecular simulations. It includes a dedicated AI node equipped with 4× NVIDIA L40S (48 GB) GPUs, offering a cost- and power-efficient alternative to A100/H100-class hardware, alongside 17 GPU nodes with a total of 132× NVIDIA RTX 4070 Ti (12 GB) GPUs optimized for scalable, general-purpose workloads. Complementing this is a high-memory CPU node (96 cores, 1.5 TB RAM), well suited for memory-intensive tasks such as multiple sequence alignments and broader CPU-bound computations. Together, this heterogeneous architecture enables efficient handling of diverse computational demands, from molecular AI to large-scale biomolecular simulations.
The facility offers support for tailored computational strategies for integrative, structure-guided research on biomolecular systems of interest.
Physical access is preferred and possible in the context of a guest scientist or guest student stay to jointly work on the tailored computational strategy for integrative, structure-guided research.
Remote access may be granted in the context of completing the tailored computational strategy for integrative, structure-guided research but usually requires physical access first.
Visit times depend on the addressed scientific question, the required service technologies, and the complexity of the system. Rough estimates are given above.
Users can access molecular simulation workflows to study soluble and (trans-)membrane proteins and their complexes with proteins, small molecules, or nucleic acids. The service employs atomistic molecular dynamics, enhanced sampling, and multiscale modeling approaches, as established in the facility, to characterize structural dynamics, interaction mechanisms, and kinetic processes. Depending on system size and complexity, expected turnaround times range from several days (standard MD) to a few weeks or months (enhanced sampling or multiscale studies).
This service provides access to workflows for calculating binding affinities and stability differences using both rigorous alchemical methods (e.g., thermodynamic integration, free energy perturbation) and efficient endpoint approaches (e.g., MM/PBSA, MM/GBSA). Users will be guided in selecting appropriate protocols balancing accuracy and computational cost, with calculations typically taking from several days to multiple weeks.
Users can address modeling of protein variants and assessment of their structural and functional impact. The service integrates structure-based modeling, rigidity theory (e.g., constraint network analysis) for stability assessment, and molecular AI methods for function prediction. Initial results can often be obtained within hours to days, while more comprehensive analyses may require up to one to two weeks.
This service supports the investigation of allosteric regulation mechanisms using a combination of molecular simulations and rigidity-based network analysis. Approaches include ensemble simulations, perturbation analyses, and identification of communication pathways to uncover long-range coupling effects. Depending on the level of detail required, analyses typically take from several days to a few weeks.
