Machine Learning-Driven Multiscale Modeling: Bridging the Scales with a Next-Generation Simulation Infrastructure.

Interdependence across time and length scales is common in biology, where atomic interactions can impact larger-scale phenomenon. Such dependence is especially true for a well-known cancer signaling pathway, where the membrane-bound RAS protein binds an effector protein called RAF. To capture the driving forces that bring RAS and RAF (represented as two domains, RBD and CRD) together on the plasma membrane, simulations with the ability to calculate atomic detail while having long time and large length- scales are needed. The Multiscale Machine-Learned Modeling Infrastructure (MuMMI) is able to resolve RAS/RAF protein-membrane interactions that identify specific lipid-protein fingerprints that enhance protein orientations viable for effector binding. MuMMI is a fully automated, ensemble-based multiscale approach connecting three resolution scales: (1) the coarsest scale is a continuum model able to simulate milliseconds of time for a 1 µm2 membrane, (2) the middle scale is a coarse-grained (CG) Martini bead model to explore protein-lipid interactions, and (3) the finest scale is an all-atom (AA) model capturing specific interactions between lipids and proteins. MuMMI dynamically couples adjacent scales in a pairwise manner using machine learning (ML). The dynamic coupling allows for better sampling of the refined scale from the adjacent coarse scale (forward) and on-the-fly feedback to improve the fidelity of the coarser scale from the adjacent refined scale (backward). MuMMI operates efficiently at any scale, from a few compute nodes to the largest supercomputers in the world, and is generalizable to simulate different systems. As computing resources continue to increase and multiscale methods continue to advance, fully automated multiscale simulations (like MuMMI) will be commonly used to address complex science questions.

The impact of pre-transplant weight loss on survival following cardiac transplantation.

BACKGROUND: Significant weight loss due to cardiac cachexia is an independent predictor of mortality in many heart failure (HF) clinical trials. The impact of significant weight loss while on the waitlist for heart transplant (HT) has yet to be studied with respect to post-transplant survival. METHODS: Adult HT recipients from 2010 to 2021 were identified in the UNOS registry. Patients who experienced an absolute weight change from the time of listing to transplant were included and classified into two groups by percent weight loss from time of listing to time of transplant using a cut-off of 10%. The primary endpoint was 1-year survival following HT. RESULTS: 5951 patients were included in the analysis, of whom 763 (13%) experienced ≥10% weight loss from the time of listing to transplant. Weight loss ≥ 10% was associated with reduced 1-year post-transplant survival (86.9% vs. 91.0%, long-rank p = .0003). Additionally, weight loss ≥ 10% was an independent predictor of 1-year mortality in a multivariable model adjusting for significant risk factors (adjusted HR 1.23, 95% CI 1.04-1.46). In secondary analyses, weight loss ≥ 10% was associated with reduced 1-year survival independent of hospitalized status at time of transplant as well as obesity status at listing (i.e., body mass index [BMI] < 30 kg/m2 and BMI ≥ 30 kg/m2 ). CONCLUSIONS: Preoperative weight loss ≥ 10% is associated with reduced survival in patients listed for HT. Nutrition interventions prior to transplant may prove beneficial in this population.

Machine learning-driven multiscale modeling reveals lipid-dependent dynamics of RAS signaling proteins.

RAS is a signaling protein associated with the cell membrane that is mutated in up to 30% of human cancers. RAS signaling has been proposed to be regulated by dynamic heterogeneity of the cell membrane. Investigating such a mechanism requires near-atomistic detail at macroscopic temporal and spatial scales, which is not possible with conventional computational or experimental techniques. We demonstrate here a multiscale simulation infrastructure that uses machine learning to create a scale-bridging ensemble of over 100,000 simulations of active wild-type KRAS on a complex, asymmetric membrane. Initialized and validated with experimental data (including a new structure of active wild-type KRAS), these simulations represent a substantial advance in the ability to characterize RAS-membrane biology. We report distinctive patterns of local lipid composition that correlate with interfacially promiscuous RAS multimerization. These lipid fingerprints are coupled to RAS dynamics, predicted to influence effector binding, and therefore may be a mechanism for regulating cell signaling cascades.

Pathogenic LMNA variants disrupt cardiac lamina-chromatin interactions and de-repress alternative fate genes.

Pathogenic mutations in LAMIN A/C (LMNA) cause abnormal nuclear structure and laminopathies. These diseases have myriad tissue-specific phenotypes, including dilated cardiomyopathy (DCM), but how LMNA mutations result in tissue-restricted disease phenotypes remains unclear. We introduced LMNA mutations from individuals with DCM into human induced pluripotent stem cells (hiPSCs) and found that hiPSC-derived cardiomyocytes, in contrast to hepatocytes or adipocytes, exhibit aberrant nuclear morphology and specific disruptions in peripheral chromatin. Disrupted regions were enriched for transcriptionally active genes and regions with lower LAMIN B1 contact frequency. The lamina-chromatin interactions disrupted in mutant cardiomyocytes were enriched for genes associated with non-myocyte lineages and correlated with higher expression of those genes. Myocardium from individuals with LMNA variants similarly showed aberrant expression of non-myocyte pathways. We propose that the lamina network safeguards cellular identity and that pathogenic LMNA variants disrupt peripheral chromatin with specific epigenetic and molecular characteristics, causing misexpression of genes normally expressed in other cell types.

Machine Learning Discovery of Computational Model Efficacy Boundaries.

Computational models are formulated in hierarchies of variable fidelity, often with no quantitative rule for defining the fidelity boundaries. We have constructed a dataset from a wide range of atomistic computational models to reveal the accuracy boundary between higher-fidelity models and a simple, lower-fidelity model. The symbolic decision boundary is discovered by optimizing a support vector machine on the data through iterative feature engineering. This data-driven approach reveals two important results: (i) a symbolic rule emerges that is independent of the algorithm, and (ii) the symbolic rule provides a deeper understanding of the fidelity boundary. Specifically, our dataset is composed of radial distribution functions from seven high-fidelity methods that cover wide ranges in the features (element, density, and temperature); high-fidelity results are compared with a simple pair-potential model to discover the nonlinear combination of the features, and the machine learning approach directly reveals the central role of atomic physics in determining accuracy.

Molecular dynamics studies of electron-ion temperature equilibration in hydrogen plasmas within the coupled-mode regime.

We use classical molecular dynamics (MD) to study electron-ion temperature equilibration in two-component plasmas in regimes for which the presence of coupled collective modes has been predicted to substantively reduce the equilibration rate. Guided by previous kinetic theory work, we examine hydrogen plasmas at a density of n=10^{26}cm^{-3}, T_{i}=10^{5}K, and 10^{7}K

A generalized Ewald decomposition for screened Coulomb interactions.

Medium-range interactions occur in a wide range of systems, including charged-particle systems with varying screening lengths. We generalize the Ewald method to charged systems described by interactions involving an arbitrary dielectric response function ϵ(𝐤). We provide an error estimate and optimize the generalization to find the break-even parameters that separate a neighbor list-only algorithm from the particle-particle particle-mesh algorithm. We examine the implications of different choices of the screening length for the computational cost of computing the dynamic structure factor. We then use our new method in molecular dynamics simulations to compute the dynamic structure factor for a model plasma system and examine the wave-dispersion properties of this system.

Ionic transport in high-energy-density matter.

Ionic transport coefficients for dense plasmas have been numerically computed using an effective Boltzmann approach. We have developed a simplified effective potential approach that yields accurate fits for all of the relevant cross sections and collision integrals. Our results have been validated with molecular-dynamics simulations for self-diffusion, interdiffusion, viscosity, and thermal conductivity. Molecular dynamics has also been used to examine the underlying assumptions of the Boltzmann approach through a categorization of behaviors of the velocity autocorrelation function in the Yukawa phase diagram. Using a velocity-dependent screening model, we examine the role of dynamical screening in transport. Implications of these results for Coulomb logarithm approaches are discussed.