Renormalized charge and dielectric effects in colloidal interactions: a numerical solution of the nonlinear Poisson-Boltzmann equation for unknown boundary conditions.

 The Derjaguin-Landau-Verwey-Overbeek (DLVO) theory, introduced more than 70 years ago, is a hallmark of colloidal particle modeling. For highly charged particles in the dilute regime, it is often supplemented by Alexander's prescription (Alexander et al. in J Chem Phys 80:5776, 1984) for using a renormalized charge. Here, we solve the problem of the interaction between two charged colloids at finite ionic strength, including dielectric mismatch effects, using an efficient numerical scheme to solve the nonlinear Poisson-Boltzmann (NPB) equation with unknown boundary conditions. Our results perfectly match the analytical predictions for the renormalized charge by Trizac and coworkers (Aubouy et al. in J Phys A 36:5835, 2003). Moreover, they allow us to reinterpret previous molecular dynamics (MD) simulation results by Kreer et al. (Phys Rev E 74:021401, 2006), rendering them now in agreement with the expected behavior. We furthermore find that the influence of polarization becomes important only when the Debye layers overlap significantly.

A coarse-grained DNA model for the prediction of current signals in DNA translocation experiments.

We present an implicit solvent coarse-grained double-stranded DNA (dsDNA) model confined to an infinite cylindrical pore that reproduces the experimentally observed current modulations of a KaCl solution at various concentrations. Our model extends previous coarse-grained and mean-field approaches by incorporating a position dependent friction term on the ions, which Kesselheim et al. [Phys. Rev. Lett. 112, 018101 (2014)] identified as an essential ingredient to correctly reproduce the experimental data of Smeets et al. [Nano Lett. 6, 89 (2006)]. Our approach reduces the computational effort by orders of magnitude compared with all-atom simulations and serves as a promising starting point for modeling the entire translocation process of dsDNA. We achieve a consistent description of the system's electrokinetics by using explicitly parameterized ions, a friction term between the DNA beads and the ions, and a lattice-Boltzmann model for the solvent.

Origin of current blockades in nanopore translocation experiments.

We present a detailed investigation of the ionic current in a cylindrical model nanopore in the absence and the presence of a double stranded DNA homopolymer. Our atomistic simulations are capable of reproducing almost exactly the experimental data obtained by Smeets et al., including notably the crossover salt concentration that yields equal current measurements in both situations. We can rule out that the observed current blockade is due to the steric exclusion of charge carriers from the DNA, since for all investigated salt concentrations the charge carrier density is higher when the DNA is present. Calculations using a mean-field electrokinetic model proposed by van Dorp et al. fail quantitatively in predicting this effect. We can relate the shortcomings of the mean-field model to a surface related molecular drag that the ions feel in the presence of the DNA. This drag is independent of the salt concentration and originates from electrostatic, hydrodynamic, and excluded volume interactions.

Hydrodynamic interactions slow down crystallization of soft colloids.

Colloidal suspensions are often argued to be an ideal model for studying phase transitions such as crystallization, as they have the advantage of tunable interactions and experimentally tractable time and length scales. Because crystallization is assumed to be unaffected by details of particle transport other than the bulk diffusion coefficient, findings are frequently argued to be transferable to pure melts without solvent. In this article, we present molecular dynamics simulations of crystallization in a suspension of colloids with Yukawa interactions which challenge this assumption. In order to investigate the role of hydrodynamic interactions mediated by the solvent, we model the solvent both implicitly and explicitly, using Langevin dynamics and the fluctuating lattice Boltzmann method, respectively. Our simulations show a significant reduction of the crystal growth velocity due to hydrodynamic interactions even at moderate hydrodynamic coupling. This slowdown is accompanied by a reduction of the width of the layering region in front of the growing crystal. Thus the dynamics of a colloidal suspension differ strongly from that of a melt, making it less useful as a model for solvent-free melts than previously thought.

Analysis of Rutherford backscattering spectra with CNN-GRU mixture density network.

Ion Beam Analysis (IBA) utilizing MeV ion beams provides valuable insights into surface elemental composition across the entire periodic table. While ion beam measurements have advanced towards high throughput for mapping applications, data analysis has lagged behind due to the challenges posed by large volumes of data and multiple detectors providing diverse analytical information. Traditional physics-based fitting algorithms for these spectra can be time-consuming and prone to local minima traps, often taking days or weeks to complete. This study presents an approach employing a Mixture Density Network (MDN) to model the posterior distribution of Elemental Depth Profiles (EDP) from input spectra. Our MDN architecture includes an encoder module (EM), leveraging a Convolutional Neural Network-Gated Recurrent Unit (CNN-GRU), and a Mixture Density Head (MDH) employing a Multi-Layer Perceptron (MLP). Validation across three datasets with varying complexities demonstrates that for simple and intermediate cases, the MDN performs comparably to the conventional automatic fitting method (Autofit). However, for more complex datasets, Autofit still outperforms the MDN. Additionally, our integrated approach, combining MDN with the automatic fit method, significantly enhances accuracy while still reducing computational time, offering a promising avenue for improved analysis in IBA.

Automatic heliostat learning for in situ concentrating solar power plant metrology with differentiable ray tracing.

Concentrating solar power plants are a clean energy source capable of competitive electricity generation even during night time, as well as the production of carbon-neutral fuels, offering a complementary role alongside photovoltaic plants. In these power plants, thousands of mirrors (heliostats) redirect sunlight onto a receiver, potentially generating temperatures exceeding 1000°C. Practically, such efficient temperatures are never attained. Several unknown, yet operationally crucial parameters, e.g., misalignment in sun-tracking and surface deformations can cause dangerous temperature spikes, necessitating high safety margins. For competitive levelized cost of energy and large-scale deployment, in-situ error measurements are an essential, yet unattained factor. To tackle this, we introduce a differentiable ray tracing machine learning approach that can derive the irradiance distribution of heliostats in a data-driven manner from a small number of calibration images already collected in most solar towers. By applying gradient-based optimization and a learning non-uniform rational B-spline heliostat model, our approach is able to determine sub-millimeter imperfections in a real-world setting and predict heliostat-specific irradiance profiles, exceeding the precision of the state-of-the-art and establishing full automatization. The new optimization pipeline enables concurrent training of physical and data-driven models, representing a pioneering effort in unifying both paradigms for concentrating solar power plants and can be a blueprint for other domains.

MISATO: machine learning dataset of protein-ligand complexes for structure-based drug discovery.

Large language models have greatly enhanced our ability to understand biology and chemistry, yet robust methods for structure-based drug discovery, quantum chemistry and structural biology are still sparse. Precise biomolecule-ligand interaction datasets are urgently needed for large language models. To address this, we present MISATO, a dataset that combines quantum mechanical properties of small molecules and associated molecular dynamics simulations of ~20,000 experimental protein-ligand complexes with extensive validation of experimental data. Starting from the existing experimental structures, semi-empirical quantum mechanics was used to systematically refine these structures. A large collection of molecular dynamics traces of protein-ligand complexes in explicit water is included, accumulating over 170 µs. We give examples of machine learning (ML) baseline models proving an improvement of accuracy by employing our data. An easy entry point for ML experts is provided to enable the next generation of drug discovery artificial intelligence models.

Providing AI expertise as an infrastructure in academia.

Artificial intelligence (AI) is proliferating and developing faster than any domain scientist can adapt. To support the scientific enterprise in the Helmholtz association, a network of AI specialists has been set up to disseminate AI expertise as an infrastructure among domain scientists. As this effort exposes an evolutionary step in science organization in Germany, this article aspires to describe our setup, goals, and motivations. We comment on past experiences, current developments, and future ideas as we bring our expertise as an infrastructure closer to scientists across our organization. We hope that this offers a brief yet insightful view of our activities as well as inspiration for other science organizations.

RNA contact prediction by data efficient deep learning.

On the path to full understanding of the structure-function relationship or even design of RNA, structure prediction would offer an intriguing complement to experimental efforts. Any deep learning on RNA structure, however, is hampered by the sparsity of labeled training data. Utilizing the limited data available, we here focus on predicting spatial adjacencies ("contact maps") as a proxy for 3D structure. Our model, BARNACLE, combines the utilization of unlabeled data through self-supervised pre-training and efficient use of the sparse labeled data through an XGBoost classifier. BARNACLE shows a considerable improvement over both the established classical baseline and a deep neural network. In order to demonstrate that our approach can be applied to tasks with similar data constraints, we show that our findings generalize to the related setting of accessible surface area prediction.

Acquisitions with random shim values enhance AI-driven NMR shimming.

Shimming is still an unavoidable, time-consuming and cumbersome burden that precedes NMR experiments, and aims to achieve a homogeneous magnetic field distribution, which is required for expressive spectroscopy measurements. This study presents multiple enhancements to AI-driven shimming. We achieve fast, quasi-iterative shimming on multiple shims simultaneously via a temporal history that combines spectra and past shim actions. Moreover, we enable efficient data collection by randomized dataset acquisition, allowing scalability to higher-order shims. Application at a low-field benchtop magnet reduces the linewidth in 87 of 100 random distortions from ∼ 4 Hz to below 1 Hz, within less than 10 NMR acquisitions. Compared to, and combined with, traditional methods, we significantly enhance both the speed and performance of shimming algorithms. In particular, AI-driven shimming needs roughly 1/3 acquisitions, and helps to avoid local minima in 96% of the cases. Our dataset and code is publicly available.

A hybrid process for chiral separation of compound-forming systems.

The resolution of chiral compound-forming systems using hybrid processes was discussed recently. The concept is of large relevance as these systems form the majority of chiral substances. In this study, a novel hybrid process is presented, which combines pertraction and subsequent preferential crystallization and is applicable for the resolution of such systems. A supported liquid membrane applied in a pertraction process provides enantiomeric enrichment. This membrane contains a solution of a chiral compound acting as a selective carrier for one of the enantiomers. Screening of a large number of liquid membranes and potential carriers using the conductor-like screening model for realistic solvation method led to the identification of several promising carriers, which were tested experimentally in several pertraction runs aiming to yield enriched (+)-(S)-mandelic acid (MA) solutions from racemic feed solutions. The most promising system consisted of tetrahydronaphthalene as liquid membrane and hydroquinine-4-methyl-2-quinolylether (HMQ) as chiral carrier achieving enantiomeric excesses of 15% in average. The successful production of (+)-(S)-MA with a purity above 96% from enriched solutions by subsequent preferential crystallization proved the applicability of the hybrid process.

Visual analysis for space-time aggregation of biomolecular simulations.

Conducting a current through a nanopore allows for the analysis of molecules inside the pore because a current modulation caused by the electrostatic properties of the passing molecules can be measured. This mechanism shows great potential for DNA sequencing, as the four different nucleotide bases induce different current modulations. We present a visualisation approach to investigate this phenomenon in our simulations of DNA within a nanopore by combining state-of-the-art molecular visualisation with vector field illustration. By spatial and temporal aggregation of the ions transported through the pore, we construct a velocity field which exhibits the induced current modulations caused by ion flux. In our interactive analysis using parametrisable three-dimensional visualisations, we encountered regions where the ion motion unexpectedly opposes the direction of the applied electric field.

Tracer diffusion in a crowded cylindrical channel.

Based on a coarse-grained model, we carry out molecular dynamics simulations to analyze the diffusion of a small tracer particle inside a cylindrical channel whose inner wall is covered with randomly grafted short polymeric chains. We observe an interesting transient subdiffusive behavior along the cylindrical axis at high attraction between the tracer and the chains, however, the long-time diffusion is always normal. This process is found to be enhanced for the case that we immobilize the grafted chains, i.e., the subdiffusive behavior sets in at an earlier time and spans over a longer time period before becoming diffusive. Even if the grafted chains are replaced with a frozen sea of repulsive, nonconnected particles in the background, a transient subdiffusion is observed. The intermediate subdiffusive behavior only disappears when the grafted chains are replaced with a mobile background sea of mutually repulsive particles. Overall, the long-time diffusion coefficient of the tracer along the cylinder axis decreases with an increase in system volume fraction, the strength of the attraction between the tracer and the background, and also on freezing the background.