Physics-Based Machine Learning Models Predict Carbon Dioxide Solubility in Chemically Reactive Deep Eutectic Solvents.

Carbon dioxide (CO2) is a detrimental greenhouse gas and is the main contributor to global warming. In addressing this environmental challenge, a promising approach emerges through the utilization of deep eutectic solvents (DESs) as an ecofriendly and sustainable medium for effective CO2 capture. Chemically reactive DESs, which form chemical bonds with the CO2, are superior to nonreactive, physically based DESs for CO2 absorption. However, there are no accurate computational models that provide accurate predictions of the CO2 solubility in chemically reactive DESs. Here, we develop machine learning (ML) models to predict the solubility of CO2 in chemically reactive DESs. As training data, we collected 214 data points for the CO2 solubility in 149 different chemically reactive DESs at different temperatures, pressures, and DES molar ratios from published work. The physics-driven input features for the ML models include σ-profile descriptors that quantify the relative probability of a molecular surface segment having a certain screening charge density and were calculated with the first-principle quantum chemical method COSMO-RS. We show here that, although COSMO-RS does not explicitly calculate chemical reaction profiles, the COSMO-RS-derived σ-profile features can be used to predict bond formation. Of the models trained, an artificial neural network (ANN) provides the most accurate CO2 solubility prediction with an average absolute relative deviation of 2.94% on the testing sets. Overall, this work provides ML models that can predict CO2 solubility precisely and thus accelerate the design and application of chemically reactive DESs.

Accurate Machine Learning for Predicting the Viscosities of Deep Eutectic Solvents.

Deep eutectic solvents (DESs) are emerging as environmentally friendly designer solvents for mass transport and heat transfer processes in industrial applications; however, the lack of accurate tools to predict and thus control their viscosities under both a range of environmental factors and formulations hinders their general application. While DESs may serve as designer solvents, with nearly unlimited combinations, this unfortunately makes it experimentally infeasible to comprehensively measure the viscosities of all DESs of potential industrial interest. To assist in the design of DESs, we have developed several new machine learning (ML) models that accurately and rapidly predict the viscosities of a diverse group of DESs at different temperatures and molar ratios using, to date, one of the most comprehensive data sets containing the properties of over 670 DESs over a wide range of temperatures (278.15-385.25 K). Three ML models, including support vector regression (SVR), feed forward neural networks (FFNNs), and categorical boosting (CatBoost), were developed to predict DES viscosity as a function of temperature and molar ratio and contrasted with multilinear and two-factor polynomial regression baselines. Quantum chemistry-based, COSMO-RS-derived sigma profile (σ-profile) features were used as inputs for the ML models. The CatBoost model is excellent at externally predicting DES viscosity, as indicated by high R2 (0.99) and low root-mean-square-error (RMSE) and average absolute relative deviations (AARD) (5.22%) values for the testing data sets, and 98% of the data points lie within the 15% of AARD deviations. Furthermore, SHapley additive explanation (SHAP) analysis was employed to interpret the ML results and rationalize the viscosity predictions. The result is an ML approach that accurately predicts viscosity and will aid in accelerating the design of appropriate DESs for industrial applications.

Drugging the entire human proteome: Are we there yet?

Each of the ∼20,000 proteins in the human proteome is a potential target for compounds that bind to it and modify its function. The 3D structures of most of these proteins are now available. Here, we discuss the prospects for using these structures to perform proteome-wide virtual HTS (VHTS). We compare physics-based (docking) and AI VHTS approaches, some of which are now being applied with large databases of compounds to thousands of targets. Although preliminary proteome-wide screens are now within our grasp, further methodological developments are expected to improve the accuracy of the results.

Atomistic Simulations of Polydisperse Lignin Melts Using Simple Polydisperse Residue Input Generator.

Understanding the physics of lignin will help rationalize its function in plant cell walls as well as aiding practical applications such as deriving biofuels and bioproducts. Here, we present SPRIG (Simple Polydisperse Residue Input Generator), a program for generating atomic-detail models of random polydisperse lignin copolymer melts i.e., the state most commonly found in nature. Using these models, we use all-atom molecular dynamics (MD) simulations to investigate the conformational and dynamic properties of polydisperse melts representative of switchgrass (Panicum virgatum L.) lignin. Polydispersity, branching and monolignol sequence are found to not affect the calculated glass transition temperature, Tg. The Flory-Huggins scaling parameter for the segmental radius of gyration is 0.42 ± 0.02, indicating that the chains exhibit statistics that lie between a globular chain and an ideal Gaussian chain. Below Tg the atomic mean squared displacements are independent of molecular weight. In contrast, above Tg, they decrease with increasing molecular weight. Therefore, a monodisperse lignin melt is a good approximation to this polydisperse lignin when only static properties are probed, whereas the molecular weight distribution needs to be considered while analyzing lignin dynamics.

Targeted inhibition of MASTL kinase activity induces apoptosis in breast cancer.

Microtubule-associated serine/threonine kinase-like (MASTL) (or Greatwall kinase (GWL)) is an important cell cycle regulating kinase that regulates the G2-M transition. Uncontrolled MASTL activity is implicated in breast cancer progression. To date, very few inhibitors have been reported against this protein. Here, structure-based computational modeling indicates that the natural product flavopiridol (FLV) binds strongly to MASTL and these results are validated using molecular dynamics simulation studies. An in vitro kinase assay reveals an EC50 (effective concentration) value of FLV to be 82.1 nM and a better IC50 compared to the positive reference compound, staurosporine. FLV is found to inhibit MASTL kinase activity, arresting the cell growth in the G1 phase and inducing apoptosis in breast cancer cells. Consistent with these results differential gene expression obtained using RNA sequencing studies, and validated by RT PCR and immunoblot analysis, indicate that MASTL inhibition induces cell cycle arrest and apoptotic-related genes. Furthermore, metastasis- and inflammation-related genes are downregulated. Thus, the deregulation of MASTL signaling pathways on targeted inhibition of its kinase activity is revealed. This study lays a strong foundation for investigating FLV as a lead compound in breast cancer therapeutics.

Comparative Assessment of Pose Prediction Accuracy in RNA-Ligand Docking.

Structure-based virtual high-throughput screening is used in early-stage drug discovery. Over the years, docking protocols and scoring functions for protein-ligand complexes have evolved to improve the accuracy in the computation of binding strengths and poses. In the past decade, RNA has also emerged as a target class for new small-molecule drugs. However, most ligand docking programs have been validated and tested for proteins and not RNA. Here, we test the docking power (pose prediction accuracy) of three state-of-the-art docking protocols on 173 RNA-small molecule crystal structures. The programs are AutoDock4 (AD4) and AutoDock Vina (Vina), which were designed for protein targets, and rDock, which was designed for both protein and nucleic acid targets. AD4 performed relatively poorly. For RNA targets for which a crystal structure of a bound ligand used to limit the docking search space is available and for which the goal is to identify new molecules for the same pocket, rDock performs slightly better than Vina, with success rates of 48% and 63%, respectively. However, in the more common type of early-stage drug discovery setting, in which no structure of a ligand-target complex is known and for which a larger search space is defined, rDock performed similarly to Vina, with a low success rate of ∼27%. Vina was found to have bias for ligands with certain physicochemical properties, whereas rDock performs similarly for all ligand properties. Thus, for projects where no ligand-protein structure already exists, Vina and rDock are both applicable. However, the relatively poor performance of all methods relative to protein-target docking illustrates a need for further methods refinement.

Water molecule ordering on the surface of an intrinsically disordered protein.

The dynamics and local structure of the hydration water on surfaces of folded proteins have been extensively investigated. However, our knowledge of the hydration of intrinsically disordered proteins (IDPs) is more limited. Here, we compare the local structure of water molecules hydrating a globular protein, lysozyme, and the intrinsically disordered N-terminal of c-Src kinase (SH4UD) using molecular dynamics simulation. The radial distributions from the protein surface of the first and the second hydration shells are similar for the folded protein and the IDP. However, water molecules in the first hydration shell of both the folded protein and the IDP are perturbed from the bulk. This perturbation involves a loss of tetrahedrality, which is, however, significantly more marked for the folded protein than the IDP. This difference arises from an increase in the first hydration shell of the IDP of the fraction of hydration water molecules interacting with oxygen. The water ordering is independent of the compactness of the IDP. In contrast, the lifetimes of water molecules in the first hydration shell increase with IDP compactness, indicating a significant impact of IDP configuration on water surface pocket kinetics, which here is linked to differential pocket volumes and polarities.

Structural asymmetry in FGF23 signaling.

Chen et al. have derived cryogenic electron microscopy (cryo-EM) structures of signaling complexes of the endocrine hormone fibroblast growth factor 23 (FGF23) with fibroblast growth factor receptor (FGFR), α-Klotho, and heparin sulfate. These structures are asymmetric, leading to questions concerning in vivo function, and will facilitate structure-based drug design to modulate FGF23 signaling.

Genetic interactions between polycystin-1 and Wwtr1 in osteoblasts define a novel mechanosensing mechanism regulating bone formation in mice.

Molecular mechanisms transducing physical forces in the bone microenvironment to regulate bone mass are poorly understood. Here, we used mouse genetics, mechanical loading, and pharmacological approaches to test the possibility that polycystin-1 and Wwtr1 have interdependent mechanosensing functions in osteoblasts. We created and compared the skeletal phenotypes of control Pkd1flox/+;Wwtr1flox/+, Pkd1Oc-cKO, Wwtr1Oc-cKO, and Pkd1/Wwtr1Oc-cKO mice to investigate genetic interactions. Consistent with an interaction between polycystins and Wwtr1 in bone in vivo, Pkd1/Wwtr1Oc-cKO mice exhibited greater reductions of BMD and periosteal MAR than either Wwtr1Oc-cKO or Pkd1Oc-cKO mice. Micro-CT 3D image analysis indicated that the reduction in bone mass was due to greater loss in both trabecular bone volume and cortical bone thickness in Pkd1/Wwtr1Oc-cKO mice compared to either Pkd1Oc-cKO or Wwtr1Oc-cKO mice. Pkd1/Wwtr1Oc-cKO mice also displayed additive reductions in mechanosensing and osteogenic gene expression profiles in bone compared to Pkd1Oc-cKO or Wwtr1Oc-cKO mice. Moreover, we found that Pkd1/Wwtr1Oc-cKO mice exhibited impaired responses to tibia mechanical loading in vivo and attenuation of load-induced mechanosensing gene expression compared to control mice. Finally, control mice treated with a small molecule mechanomimetic, MS2 that activates the polycystin complex resulted in marked increases in femoral BMD and periosteal MAR compared to vehicle control. In contrast, Pkd1/Wwtr1Oc-cKO mice were resistant to the anabolic effects of MS2. These findings suggest that PC1 and Wwtr1 form an anabolic mechanotransduction signaling complex that mediates mechanical loading responses and serves as a potential novel therapeutic target for treating osteoporosis.

Structure-guided design of a broadly cross-reactive multivalent group a streptococcal vaccine.

The M protein of group A streptococci (Strep A) is a major virulence determinant and protective antigen. The N-terminal region of the M protein is variable in sequence, defines the M/emm type, and contains epitopes that elicit opsonic antibodies that protect animals from challenge infections. Although there are >200 M types of Strep A, there is now evidence that structurally related M proteins can be grouped into clusters and that immunity may be cluster-specific in addition to M type-specific. This observation has led to recent studies of structure-based design of multivalent M peptide vaccines to select peptides predicted to cross-react with heterologous M types to improve vaccine coverage. In the current study, we have applied a refined series of peptide structural algorithms to predict immunological cross-reactivity among 117 N-terminal M peptides representing the most prevalent M types of Strep A. Based on the results of the structural analyses, in combination with global M type prevalence data, we constructed a 32-valent vaccine containing 19 cross-reactive vaccine candidates predicted to cross-react with 37 heterologous M peptides to which were added 13 type-specific M peptides. The 4-protein recombinant vaccine was immunogenic in rabbits and elicited significant levels of antibodies against 31/32 (97%) vaccine peptides and 28/37 (76%) peptides predicted to cross-react. The vaccine antisera also promoted opsonophagocytic killing of vaccine and cross-reactive M types of Strep A. Based on a recent analysis of M type prevalence of Strep A, the potential global coverage of the 32-valent vaccine is âˆ¼90%, ranging from 68% in Africa to 95% in North America. Our results indicate the utility of structure-based design that may be applied to future studies of broadly protective M peptide vaccines.

Genetic interactions between Polycystin-1 and TAZ in osteoblasts define a novel mechanosensing mechanism regulating bone formation in mice.

Molecular mechanisms transducing physical forces in the bone microenvironment to regulate bone mass are poorly understood. Here, we used mouse genetics, mechanical loading, and pharmacological approaches to test the possibility that polycystin-1 and TAZ have interdependent mechanosensing functions in osteoblasts. We created and compared the skeletal phenotypes of control Pkd1flox/+;TAZflox/+, single Pkd1Oc-cKO, single TAZOc-cKO, and double Pkd1/TAZOc-cKO mice to investigate genetic interactions. Consistent with an interaction between polycystins and TAZ in bone in vivo, double Pkd1/TAZOc-cKO mice exhibited greater reductions of BMD and periosteal MAR than either single TAZOc-cKO or Pkd1Oc-cKO mice. Micro-CT 3D image analysis indicated that the reduction in bone mass was due to greater loss in both trabecular bone volume and cortical bone thickness in double Pkd1/TAZOc-cKO mice compared to either single Pkd1Oc-cKO or TAZOc-cKO mice. Double Pkd1/TAZOc-cKO mice also displayed additive reductions in mechanosensing and osteogenic gene expression profiles in bone compared to single Pkd1Oc-cKO or TAZOc-cKO mice. Moreover, we found that double Pkd1/TAZOc-cKO mice exhibited impaired responses to tibia mechanical loading in vivo and attenuation of load-induced mechanosensing gene expression compared to control mice. Finally, control mice treated with a small molecule mechanomimetic MS2 had marked increases in femoral BMD and periosteal MAR compared to vehicle control. In contrast, double Pkd1/TAZOc-cKO mice were resistant to the anabolic effects of MS2 that activates the polycystin signaling complex. These findings suggest that PC1 and TAZ form an anabolic mechanotransduction signaling complex that responds to mechanical loading and serve as a potential novel therapeutic target for treating osteoporosis.

Predictive understanding of the surface tension and velocity of sound in ionic liquids using machine learning.

Knowledge of the physical properties of ionic liquids (ILs), such as the surface tension and speed of sound, is important for both industrial and research applications. Unfortunately, technical challenges and costs limit exhaustive experimental screening efforts of ILs for these critical properties. Previous work has demonstrated that the use of quantum-mechanics-based thermochemical property prediction tools, such as the conductor-like screening model for real solvents, when combined with machine learning (ML) approaches, may provide an alternative pathway to guide the rapid screening and design of ILs for desired physiochemical properties. However, the question of which machine-learning approaches are most appropriate remains. In the present study, we examine how different ML architectures, ranging from tree-based approaches to feed-forward artificial neural networks, perform in generating nonlinear multivariate quantitative structure-property relationship models for the prediction of the temperature- and pressure-dependent surface tension of and speed of sound in ILs over a wide range of surface tensions (16.9-76.2 mN/m) and speeds of sound (1009.7-1992 m/s). The ML models are further interrogated using the powerful interpretation method, shapley additive explanations. We find that several different ML models provide high accuracy, according to traditional statistical metrics. The decision tree-based approaches appear to be the most accurate and precise, with extreme gradient-boosting trees and gradient-boosting trees being the best performers. However, our results also indicate that the promise of using machine-learning to gain deep insights into the underlying physics driving structure-property relationships in ILs may still be somewhat premature.

HLA-Clus: HLA class I clustering based on 3D structure.

BACKGROUND: In a previous paper, we classified populated HLA class I alleles into supertypes and subtypes based on the similarity of 3D landscape of peptide binding grooves, using newly defined structure distance metric and hierarchical clustering approach. Compared to other approaches, our method achieves higher correlation with peptide binding specificity, intra-cluster similarity (cohesion), and robustness. Here we introduce HLA-Clus, a Python package for clustering HLA Class I alleles using the method we developed recently and describe additional features including a new nearest neighbor clustering method that facilitates clustering based on user-defined criteria. RESULTS: The HLA-Clus pipeline includes three stages: First, HLA Class I structural models are coarse grained and transformed into clouds of labeled points. Second, similarities between alleles are determined using a newly defined structure distance metric that accounts for spatial and physicochemical similarities. Finally, alleles are clustered via hierarchical or nearest-neighbor approaches. We also interfaced HLA-Clus with the peptide:HLA affinity predictor MHCnuggets. By using the nearest neighbor clustering method to select optimal allele-specific deep learning models in MHCnuggets, the average accuracy of peptide binding prediction of rare alleles was improved. CONCLUSIONS: The HLA-Clus package offers a solution for characterizing the peptide binding specificities of a large number of HLA alleles. This method can be applied in HLA functional studies, such as the development of peptide affinity predictors, disease association studies, and HLA matching for grafting. HLA-Clus is freely available at our GitHub repository ( https://github.com/yshen25/HLA-Clus ).

SCOMAP-XD: atomistic deuterium contrast matching for small-angle neutron scattering in biology.

The contrast-variation method in small-angle neutron scattering (SANS) is a uniquely powerful technique for determining the structure of individual components in biomolecular systems containing regions of different neutron scattering length density ρ. By altering the ρ of the target solute and the solvent through judicious incorporation of deuterium, the scattering of desired solute features can be highlighted. Most contrast-variation methods focus on highlighting specific bulk solute elements, but not on how the scattering at specific scattering vectors q, which are associated with specific structural distances, changes with contrast. Indeed, many systems exhibit q-dependent contrast effects. Here, a method is presented for calculating both bulk contrast-match points and q-dependent contrast using 3D models with explicit solute and solvent atoms and SASSENA, an explicit-atom SANS calculator. The method calculates the bulk contrast-match points within 2.4% solvent D2O accuracy for test protein-nucleic acid and lipid nanodisc systems. The method incorporates a general model for the incorporation of deuterium at non-exchangeable sites that was derived by performing mass spectrometry on green fluorescent protein. The method also decomposes the scattering profile into its component parts and identifies structural features that change with contrast. The method is readily applicable to a variety of systems, will expand the understanding of q-dependent contrast matching and will aid in the optimization of next-generation neutron scattering experiments.

Dissecting Phenotype from Genotype with Clinical Isolates of SARS-CoV-2 First Wave Variants.

The emergence and availability of closely related clinical isolates of SARS-CoV-2 offers a unique opportunity to identify novel nonsynonymous mutations that may impact phenotype. Global sequencing efforts show that SARS-CoV-2 variants have emerged and then been replaced since the beginning of the pandemic, yet we have limited information regarding the breadth of variant-specific host responses. Using primary cell cultures and the K18-hACE2 mouse, we investigated the replication, innate immune response, and pathology of closely related, clinical variants circulating during the first wave of the pandemic. Mathematical modeling of the lung viral replication of four clinical isolates showed a dichotomy between two B.1. isolates with significantly faster and slower infected cell clearance rates, respectively. While isolates induced several common immune host responses to infection, one B.1 isolate was unique in the promotion of eosinophil-associated proteins IL-5 and CCL11. Moreover, its mortality rate was significantly slower. Lung microscopic histopathology suggested further phenotypic divergence among the five isolates showing three distinct sets of phenotypes: (i) consolidation, alveolar hemorrhage, and inflammation, (ii) interstitial inflammation/septal thickening and peribronchiolar/perivascular lymphoid cells, and (iii) consolidation, alveolar involvement, and endothelial hypertrophy/margination. Together these findings show divergence in the phenotypic outcomes of these clinical isolates and reveal the potential importance of nonsynonymous mutations in nsp2 and ORF8.

SARS-CoV2 billion-compound docking.

This dataset contains ligand conformations and docking scores for 1.4 billion molecules docked against 6 structural targets from SARS-CoV2, representing 5 unique proteins: MPro, NSP15, PLPro, RDRP, and the Spike protein. Docking was carried out using the AutoDock-GPU platform on the Summit supercomputer and Google Cloud. The docking procedure employed the Solis Wets search method to generate 20 independent ligand binding poses per compound. Each compound geometry was scored using the AutoDock free energy estimate, and rescored using RFScore v3 and DUD-E machine-learned rescoring models. Input protein structures are included, suitable for use by AutoDock-GPU and other docking programs. As the result of an exceptionally large docking campaign, this dataset represents a valuable resource for discovering trends across small molecule and protein binding sites, training AI models, and comparing to inhibitor compounds targeting SARS-CoV-2. The work also gives an example of how to organize and process data from ultra-large docking screens.

Disordered Domain Shifts the Conformational Ensemble of the Folded Regulatory Domain of the Multidomain Oncoprotein c-Src.

c-Src kinase is a multidomain non-receptor tyrosine kinase that aberrantly phosphorylates several signaling proteins in cancers. Although the structural properties of the regulatory domains (SH3-SH2) and the catalytic kinase domain have been extensively characterized, there is less knowledge about the N-terminal disordered region (SH4UD) and its interactions with the other c-Src domains. Here, we used domain-selective isotopic labeling combined with the small-angle neutron scattering contrast matching technique to study SH4UD interactions with SH3-SH2. Our results show that in the presence of SH4UD, the radius of gyration (Rg) of SH3-SH2 increases, indicating that it has a more extended conformation. Hamiltonian replica exchange molecular dynamics simulations provide a detailed molecular description of the structural changes in SH4UD-SH3-SH2 and show that the regulatory loops of SH3 undergo significant conformational changes in the presence of SH4UD, while SH2 remains largely unchanged. Overall, this study highlights how a disordered region can drive a folded region of a multidomain protein to become flexible, which may be important for allosteric interactions with binding partners. This may help in the design of therapeutic interventions that target the regulatory domains of this important family of kinases.

HLA Class I Supertype Classification Based on Structural Similarity.

HLA class I proteins, a critical component in adaptive immunity, bind and present intracellular Ags to CD8+ T cells. The extreme polymorphism of HLA genes and associated peptide binding specificities leads to challenges in various endeavors, including neoantigen vaccine development, disease association studies, and HLA typing. Supertype classification, defined by clustering functionally similar HLA alleles, has proven helpful in reducing the complexity of distinguishing alleles. However, determining supertypes via experiments is impractical, and current in silico classification methods exhibit limitations in stability and functional relevance. In this study, by incorporating three-dimensional structures we present a method for classifying HLA class I molecules with improved breadth, accuracy, stability, and flexibility. Critical for these advances is our finding that structural similarity highly correlates with peptide binding specificity. The new classification should be broadly useful in peptide-based vaccine development and HLA-disease association studies.

Speed vs Accuracy: Effect on Ligand Pose Accuracy of Varying Box Size and Exhaustiveness in AutoDock Vina.

Structure-based virtual high-throughput screening involves docking chemical libraries to targets of interest. A parameter pertinent to the accuracy of the resulting pose is the root mean square deviation (RMSD) from a known crystallographic structure, i. e., the 'docking power'. Here, using a popular algorithm, Autodock Vina, as a model program, we evaluate the effects of varying two common docking parameters: the box size (the size of docking search space) and the exhaustiveness of the global search (the number of independent runs starting from random ligand conformations) on the RMSD from the PDBbind v2017 refined dataset of experimental protein-ligand complexes. Although it is clear that exhaustiveness is an important parameter, there is wide variation in the values used, with variation between 1 and >100. We, therefore, evaluated a combination of cubic boxes of different sizes and five exhaustiveness values (1, 8, 25, 50, 75, 100) within the range of those commonly adopted. The results show that the default exhaustiveness value of 8 performs well overall for most box sizes. In contrast, for all box sizes, but particularly for large boxes, an exhaustiveness value of 1 led to significantly higher median RMSD (mRMSD) values. The docking power was slightly improved with an exhaustiveness of 25, but the mRMSD changes little with values higher than 25. Therefore, although low exhaustiveness is computationally faster, the results are more likely to be far from reality, and, conversely, values >25 led to little improvement at the expense of computational resources. Overall, we recommend users to use at least the default exhaustiveness value of 8 for virtual screening calculations.