High-throughput Molecular Simulations Research Articles

High-throughput molecular simulations of SARS-CoV-2 receptor binding domain mutants quantify correlations between dynamic fluctuations and protein expression.

Prediction of protein fitness from computational modeling is an area of active research in rational protein design. Here, we investigated whether protein fluctuations computed from molecular dynamics simulations can be used to predict the expression levels of SARS-CoV-2 receptor binding domain (RBD) mutants determined in the deep mutational scanning experiment of Starr et al. [Science (New York, N.Y.) 2022, 377, 420] Specifically, we performed more than 0.7 milliseconds of molecular dynamics (MD) simulations of 557 mutant RBDs in triplicate to achieve statistical significance under various simulation conditions. Our results show modest but significant anticorrelation in the range [-0.4, -0.3] between expression and RBD protein flexibility. A simple linear regression machine learning model achieved correlation coefficients in the range [0.7, 0.8], thus outperforming MD-based models, but required about 25 mutations at each residue position for training.

Combining machine learning and molecular simulation to explore MOF materials that contribute to CF4/N2 separation

High-throughput molecular simulations and machine learning algorithms have been widely used to identify promising metal–organic frameworks for gas separation. However, most studies are limited to screening high-performance materials from existing databases, which fails to fully utilize the predictive function of machine learning. This paper combines genetic algorithms and high-throughput screening to deeply mine anion-pillared MOF (APMOF) performance feature relationships and predict high-performance materials that are not in the database. Considering the actual industrial conditions (CF4/N2 = 10/90), we chose the ratio of CF4:N2 = 1:9 to simulate the adsorption separation of gas mixtures of MOF materials at a temperature of 298 K and a pressure of 1 bar. First, the CF4/N2 separation properties of MOFs in the APMOF library were obtained based on molecular simulations. Then, the filtered data were coded according to the method of “building block classification − structural interval categorization”. Then, the machine learning algorithm was used for model training to obtain a high-precision model. Finally, the tangent adaptive genetic algorithm was used to recombine the genes of the materials, and the new MOF materials were successfully reverse-engineered. The study found that the pore-limit diameter of APMOFs is most conducive to the separation of CF4/N2 by MOFs when the pore-limit diameter of APMOFs is within two times the molecular dynamics diameter of CF4. 134 MOF materials were predicted to have CF4/N2 selectivities exceeding 46.30. The use of organic ligands such as 4,4′-bipyridyl or 1,4-bis(4-pyridyl)benzene (bpb) increases the likelihood of these materials being high-performance for CF4/N2 separation. The combination of computational screening methods and machine learning can expedite the design and development of new high-performance MOFs. This approach can also provide guidance for the synthetic direction and pattern of materials, which can assist experimentalists in their synthesis.

Combining computational screening and machine learning to explore MOFs and COFs for methane purification

Metal-organic frameworks (MOFs) and covalent organic frameworks (COFs) have great potential to be used as porous adsorbents and membranes to achieve high-performance methane purification. Although the continuous increase in the number and diversity of MOFs and COFs is a great opportunity for the discovery of novel adsorbents and membranes with superior performances, evaluating such a vast number of materials in the quickest and most effective manner requires the development of computational approaches. High-throughput computational screening based on molecular simulations has been extensively used to identify the most promising MOFs and COFs for methane purification. However, the enormous and ever-growing material space necessitates more efficient approaches in terms of time and effort. Combining data science with molecular simulations has recently accelerated the discovery of optimal MOF and COF materials for methane purification and revealed the hidden structure–performance relationships. In this perspective, we highlighted the recent developments in combining high-throughput molecular simulations and machine learning to accurately identify the most promising MOF and COF adsorbents and membranes among thousands of candidates for separating methane from other gases including acetylene, carbon dioxide, helium, hydrogen, and nitrogen. After providing a brief overview of the topic, we reviewed the pioneering contributions in the field and discussed the current opportunities and challenges that we need to direct our efforts for the design and discovery of adsorbent and membrane materials.

Efficient Exploration of Chemical Compound Space Using Active Learning for Prediction of Thermodynamic Properties of Alkane Molecules.

We introduce an exploratory active learning (AL) algorithm using Gaussian process regression and marginalized graph kernel (GPR-MGK) to sample chemical compound space (CCS) at minimal cost. Targeting 251,728 enumerated alkane molecules with 4-19 carbon atoms, we applied the AL algorithm to select a diverse and representative set of molecules and then conducted high-throughput molecular simulations on these selected molecules. To demonstrate the power of the AL algorithm, we built directed message-passing neural networks (D-MPNN) using simulation data as the training set to predict liquid densities, heat capacities, and vaporization enthalpies of the CCS. Validations show that D-MPNN models built on the smallest training set considered in this work, which consists of 313 molecules or 0.124% of the original CCS, predict the properties with R2 > 0.99 against the computational data and R2 > 0.94 against the experimental data. The advantage of the presented AL algorithm is that the predicted uncertainty of GPR depends on only the molecular structures, which renders it compatible with high-throughput data generation.

Efficient Generation of Large Collections of Metal-Organic Framework Structures Containing Well-Defined Point Defects.

High-throughput molecular simulations of metal-organic frameworks (MOFs) are a useful complement to experiments to identify candidates for chemical separation and storage. All previous efforts of this kind have used simulations in which MOFs are approximated as defect-free. We introduce a tool to readily generate missing-linker defects in MOFs and demonstrate this tool with a collection of 507 defective MOFs. We introduce the concept of the maximum possible defect concentration; at higher defect concentrations, deviations from the defect-free crystal structure would be readily evident experimentally. We studied the impact of defects on molecular adsorption as a function of defect concentrations. Defects have a slightly negative or negligible influence on adsorption at low pressures for ethene, ethane, and CO2 but a strong positive influence for methanol due to hydrogen bonding with defects. Defective structures tend to have loadings slightly higher than those of defect-free structures for all adsorbates at elevated pressures.

Constraints on Knot Insertion, Not Internal Jamming, Control Polycatenane Translocation Dynamics through Crystalline Pores.

The translocation of polymers through pores and channels is an archetypal process in biology and is widely studied and exploited for applications in bio- and nanotechnology. In recent times, the translocation of polymers of various different topologies has been studied both experimentally and by computer simulation. However, in some cases, a clear understanding of the precise mechanisms that drive their translocation dynamics can be challenging to derive. Experimental methods are able to provide statistical details of polymer translocation, but computer simulations are uniquely placed to uncover a finer level of mechanistic understanding. In this work, we use high-throughput molecular simulations to reveal the importance that knot insertion rates play in controlling translocation dynamics in the small pore limit, where unexpected nonpower law behavior emerges. This work both provides new predictive understanding of polycatenane translocation and shows the importance of carefully considering the role of the definition of translocation itself.

A Self-Evolutionary Methodology for Reverse Design of Novel MOFs.

A combination of high-throughput molecular simulation and machine learning (ML) algorithms has been widely adopted to seek promising metal-organic frameworks (MOFs) as energy gas carriers. However, the currently reported studies are mainly limited to extracting top performers from existing databases, not fully unleashing the ML capabilities for intelligently predicting novel structures with better performance. Herein, an efficient self-evolutionary methodology was proposed for searching high-performance MOFs that are unstructured in the origin database, in which a Tangent Adaptive Genetic Algorithm (TAGA) was newly put forward for structural evolution and the high-precision ML model of eXtreme Gradient Boosting (XGBoost) was employed as the fitness function. By taking CH4 storage in MOFs at room temperature as a showcase and using the database of 51,163 hMOFs, the TAGA-XGBoost coupling strategy rapidly suggested a certain number of possible combinations of the building blocks to form new structures with gravimetric storage capacity (35 bar) and volumetric working capacity (65-5.8 bar) higher than the best materials in the original database. The structures of some promising MOFs successfully used the finally optimized material genes for the two application conditions, and their performances were also confirmed by subsequent molecular simulations. The best materials can respectively reach a storage amount of 580 cm3(STP)/g at 35 bar and a working capacity of 218 cm3(STP)/cm3 between 65 and 5.8 bar. An analysis of the top 100 materials predicted from our method revealed that the choice of organic linkers has a systematic effect on the storage performance of MOFs. It might be believed that the proposed methodology offers an opportunity to expedite the discovery of unprecedented materials for other practical applications.

High-Throughput Computational Exploration of MOFs with Open Cu Sites for Adsorptive Separation of Hydrogen Isotopes.

Effective separation of hydrogen isotopes still remains one of the extremely challenging tasks in industry. Compared to the present methods that are energy- and cost-intensive, quantum sieving technology based on nanostructured materials offers a more efficient alternative approach, where metal-organic frameworks (MOFs) featuring open metal sites (OMS) can serve as an ideal platform. Herein, a combination of periodic density functional theory (DFT) with dispersive correction and high-throughput molecular simulation was employed from thermodynamic viewpoints to explore the D2/H2 separation properties of 929 experimental MOFs bearing a copper-paddlewheel unit. The DFT calculations showed that there is a negligible rotational energy barrier for the molecule adsorbed at the OMS, and the movement of the Cu atoms along the Cu-Cu axis vector almost has no influence on the interaction energy. On the basis of the DFT results, a new force field with a proposed cutoff scheme was developed to accurately describe the strong isotope-OMS interaction. Under practical conditions (40 K and 1.0 bar), large-scale computational material screening demonstrated that the OMS interaction plays a more important role in highly selective materials and ignoring such interactions can lead to completely wrong identification of the most promising materials. Using the adsorption selectivity and adsorbent performance score as evaluation metrics, this work demonstrated that the materials with sql topology notably outperform many benchmark adsorbents reported so far.

In silico screening and experimental study of anion-pillared metal-organic frameworks for hydrogen isotope separation

Deuterium is a key energy source for fusion reaction and widely used in various application fields. Nevertheless, the separation of deuterium from its isotopic mixture with nearly same physicochemical properties is one of the greatest challenging tasks in industry. Compared with cryogenic distillation, quantum sieving effect based on nanoporous materials has been proposed as a promising strategy for isotopes separation, in which metal-organic frameworks (MOFs) are considered as an ideal platform because of their structural diversity and tailorable functionality. In this work, high-throughput molecular simulations were performed to identify optimal structural features of high-performance materials from a recently-constructed database of anion-pillared MOFs. As a proof-of-concept, the identified material SIFSIX-18-Cd was prepared using a proposed direct mixing method, which can achieve a goal of quickly preparation of such material on a large scale. The experimental pure-component adsorption isotherms of D2 and H2 confirmed the preferential adsorption of SIFSIX-18-Cd for D2 over H2 due to the quantum sieving effect. The advanced cryogenic thermal desorption spectroscopy (ACTDS) measurements further showed that the enrichment factor (EF) of an equimolar D2/H2 mixture in SIFSIX-18-Cd can reach 5.1 at 30 K and 1.0 bar, which is higher than the current cryogenic distillation method (1.5 at 24 K). Along with the proposed efficient preparation method, the high hydrophobic property of SIFSIX-18-Cd also demonstrated a broad industrial prospect of this material for hydrogen isotope separation.

Design of Organic Electronic Materials With a Goal-Directed Generative Model Powered by Deep Neural Networks and High-Throughput Molecular Simulations.

In recent years, generative machine learning approaches have attracted significant attention as an enabling approach for designing novel molecular materials with minimal design bias and thereby realizing more directed design for a specific materials property space. Further, data-driven approaches have emerged as a new tool to accelerate the development of novel organic electronic materials for organic light-emitting diode (OLED) applications. We demonstrate and validate a goal-directed generative machine learning framework based on a recurrent neural network (RNN) deep reinforcement learning approach for the design of hole transporting OLED materials. These large-scale molecular simulations also demonstrate a rapid, cost-effective method to identify new materials in OLEDs while also enabling expansion into many other verticals such as catalyst design, aerospace, life science, and petrochemicals.

Prediction of methane storage in covalent organic frameworks using big-data-mining approach

Abstract A combination of computational materials screening and machine learning (ML) technique is being adopted as a popular approach to study various materials toward application of interest. In this work, we began with high-throughput molecular simulations to calculate the methane storage (6.5 MPa) and deliverable (6.5-0.58 MPa) capacities of 404,460 covalent organic frameworks (COFs) at 298 K. Then, the full data sets with 23 features were randomly split into training and test sets in a ratio of 20:80, which were applied to evaluate the prediction abilities of several ML algorithms, including gradient boosting decision tree (GBDT), neural network (NN), support vector machine (SVM), random forest (RF) and decision tree (DT). The results indicate that the RF model has the highest prediction accuracy, which was further employed to reduce the dimension of features space and quantitatively analyze the relative importance of each feature value. The binary classification predictors built using the features with the highest influence weight can give a successful identification of top-performing candidates from the test set containing 323,168 COFs with an accuracy exceeding 96%. The deliverable capacities of the identified COFs were found to outperform those reported so far for various adsorbents. The findings may provide a useful guidance for the design and synthesis of new high-performance materials for methane storage application.

A Data Mining Approach to Investigate the Carbon Nanotubes Mechanical Properties via High-Throughput Molecular Simulation

The relationship of geometrical properties and mechanical properties of carbon nanotubes (CNTs) was investigated by using high-throughput molecular simulation. Geometrical properties such as diameter, number of walls, chirality, and crosslink density were considered. As a key factor in determining the mechanical properties of composites reinforced with CNTs, nominal tensile strength is the focus in this study, which can be calculated by fracture force divided by the full cross-sectional area including the hollow core and the wall thickness. The fracture mode, nominal tensile strength, and nominal Young’s modulus under the condition of CNTs outermost tube loading axial tensile test were evaluated. Three types of fracture modes led by different crosslink densities of CNTs were obtained. By data-mining through large amounts of datasets, we showed that CNTs with small diameter, large number of walls, and crosslinks between walls can have high nominal tensile strength. We demonstrated that zigzag-type CNTs with crosslink density of approximately 1.5% - 2.5%, armchair-type CNTs with crosslink density of approximately 3% - 4% can help improve the load transfer from the outer tube to the inner tube the most.

The development of a comprehensive toolbox based on multi-level, high-throughput screening of MOFs for CO/N2 separations.

The separation of CO/N2 mixtures is a challenging problem in the petrochemical sector due to the very similar physical properties of these two molecules, such as size, molecular weight and boiling point. To solve this and other challenging gas separations, one requires a holistic approach. The complexity of a screening exercise for adsorption-based separations arises from the multitude of existing porous materials, including metal–organic frameworks. Besides, the multivariate nature of the performance criteria that needs to be considered when designing an optimal adsorbent and a separation process – i.e. an optimal material requires fulfillment of several criteria simultaneously – makes the screening challenging. To address this, we have developed a multi-scale approach combining high-throughput molecular simulation screening, data mining and advanced visualization, as well as process system modelling, backed up by experimental validation. We have applied our recent advances in the engineering of porous materials' morphology to develop advanced monolithic structures. These conformed, shaped monoliths can be used readily in industrial applications, bringing a valuable strategy for the development of advanced materials. This toolbox is flexible enough to be applied to multiple adsorption-based gas separation applications.

High-throughput molecular simulations reveal the origin of ion free energy barriers in graphene oxide membranes.

Graphene oxide (GO) membranes are highly touted as materials for contemporary separation challenges including desalination, yet understanding of the interplay between their structure and salt rejection is limited. K+ ion permeation through hydrated GO membranes was investigated by combining structurally realistic molecular models and high-throughput molecular dynamics simulations. We show that it is essential to consider the complex GO microstructure to quantitatively reproduce experimentally-derived free energy barriers to K+ permeation for membranes with various interlayer distances less than 1.3 nm. This finding confirms the non-uniformity of GO nanopores and the necessity of the high-throughput approach for this class of material. The large barriers arise due to significant dehydration of K+ inside the membrane, which can have as few as 3 coordinated water molecules, compared to 7 in bulk solution. Thus, even if the membranes have an average pore size larger than the ion's hydrated diameter, the significant presence of pores whose size is smaller than the hydrated diameter creates bottlenecks for the permeation process.

CO2 separation from flue gas mixture using [BMIM][BF4]/MOF composites: Linking high-throughput computational screening with experiments

In this study, we combined experiments with high-throughput molecular simulation methods to unlock CO2/N2 separation performances of 1085 different types of ionic liquid (IL)/metal organic framework (MOF) composites. We first validated the accuracy of our proposed computational methodology by synthesizing and characterizing three different IL/MOF composites composed of [BMIM][BF4] (1-n-butyl-3-methylimidazolium tetrafluoroborate) and by comparing their experimental CO2, N2 adsorption and separation performances with the simulation results. Motivated from the good agreement between experiments and simulations, we performed a high-throughput computational screening of 1085 different types of MOFs and their [BMIM][BF4]-incorporated counterparts to compute adsorption of CO2/N2 mixture in each material. Adsorbent performance evaluation metrics of [BMIM][BF4]/MOF composites including selectivity, working capacity, adsorbent performance score, and regenerability were calculated and compared with those of pristine MOFs to assess adsorption-based CO2/N2 separation performance limits of the composite materials. Our results revealed that [BMIM][BF4] incorporation remarkably increases CO2 selectivity, CO2 working capacity, and adsorption performance score of very large numbers of MOFs, resulting in excellent adsorbent candidates for separation of flue gas mixture. Analysis of the structure-performance relations showed that composites having narrow pore sizes, low porosities, and high IL loadings offer high CO2/N2 selectivities. These results will be useful in guiding and accelerating the design and development of new IL/MOF composites having exceptional CO2 capture performances from flue gas mixtures.

Computational Screening of Covalent Organic Frameworks for Hydrogen Storage

Covalent Organic Frameworks (COFs) have been considered as promising materials for gas storage applications due to their highly porous structures and tunable characteristics. In this work, high-throughput molecular simulations were performed to screen the recent Computation-Ready Experimental COF Database (CoRE-COF) for H2 storage a first time in the literature. Predictions for H2 uptakes were first compared with the experimental data of several COFs. Motivated from the good agreement between simulations and experiments, we performed Grand Canonical Monte Carlo (GCMC) simulations to compute volumetric H2 uptakes of 296 COFs at various temperatures and pressures and identified the best candidates which exhibit a superior performance for H2 storage. COFs outperformed several well-known MOFs such as HKUST-1, NU-125, NU-1000 series, NOTT-112 and UiO-67 at 100bar/77K adsorption and 5bar/160K desorption conditions. We also examined the effect of Feynman-Hibbs correction on simulated H2 isotherms and H2 working capacities of COFs to consider quantum effects at low temperatures. Results showed that the Feynman-Hibbs corrections do not affect the ranking of materials based on H2 working capacities, but slightly affect the predictions of H2 adsorption isotherms. We finally examined the structure-performance relations and showed that density and porosity are highly correlated with the volumetric H2 working capacities of COFs. Results of this study will be highly useful in guiding future research and focusing experimental efforts on the best COF adsorbents identified in this study.

It’s an Interesting MOF, but Is It Stable?

The mechanical stability of metal-organic framework (MOF) structures is key to their functionality. In this issue of Matter, Moghadam et al. use high-throughput simulations and machine learning to derive structure-mechanical properties and predict bulk moduli of MOFs from a few physical parameters.

Structure-Mechanical Stability Relations of Metal-Organic Frameworks via Machine Learning

Summary Assessing the mechanical stability of metal-organic frameworks (MOFs) is critical to bring these materials to any application. Here, we derive the first interactive map of the structure-mechanical landscape of MOFs by performing a multi-level computational analysis. First, we used high-throughput molecular simulations for 3,385 MOFs containing 41 distinct network topologies. Second, we developed a freely available machine-learning algorithm to automatically predict the mechanical properties of MOFs. For distinct regions of the high-throughput space, in-depth analysis based on in operando molecular dynamics simulations reveals the loss-of-crystallinity pressure within a given topology. The overarching mechanical screening approach presented here reveals the sensitivity on structural parameters such as topology, coordination characteristics and the nature of the building blocks, and paves the way for computational as well as experimental researchers to assess and design MOFs with enhanced mechanical stability to accelerate the translation of MOFs to industrial applications.

High-Throughput Molecular Simulations of Metal Organic Frameworks for CO2 Separation: Opportunities and Challenges

Metal organic frameworks (MOFs) have emerged as great alternatives to traditional nanoporous materials for CO2 separation applications. MOFs are porous materials that are formed by self-assembly of transition metals and organic ligands. The most important advantage of MOFs over well-known porous materials is the possibility to generate multiple materials with varying structural properties and chemical functionalities by changing the combination of metal centers and organic linkers during the synthesis. This leads to a large diversity of materials with various pore sizes and shapes that can be efficiently used for CO2 separations. Since the number of synthesized MOFs has already reached to several thousand, experimental investigation of each MOF at the lab-scale is not practical. High-throughput computational screening of MOFs is a great opportunity to identify the best materials for CO2 separation and to gain molecular-level insights into the structure-performance relationships. This type of knowledge can be used to design new materials with the desired structural features that can lead to extraordinarily high CO2 selectivities. In this mini-review, we focused on developments in high-throughput molecular simulations of MOFs for CO2 separations. After reviewing the current studies on this topic, we discussed the opportunities and challenges in the field and addressed the potential future developments.

Effects of Force Field Selection on the Computational Ranking of MOFs for CO2 Separations.

Metal–organic frameworks (MOFs) have been considered as highly promising materials for adsorption-based CO2 separations. The number of synthesized MOFs has been increasing very rapidly. High-throughput molecular simulations are very useful to screen large numbers of MOFs in order to identify the most promising adsorbents prior to extensive experimental studies. Results of molecular simulations depend on the force field used to define the interactions between gas molecules and MOFs. Choosing the appropriate force field for MOFs is essential to make reliable predictions about the materials’ performance. In this work, we performed two sets of molecular simulations using the two widely used generic force fields, Dreiding and UFF, and obtained adsorption data of CO2/H2, CO2/N2, and CO2/CH4 mixtures in 100 different MOF structures. Using this adsorption data, several adsorbent evaluation metrics including selectivity, working capacity, sorbent selection parameter, and percent regenerability were computed for each MOF. MOFs were then ranked based on these evaluation metrics, and top performing materials were identified. We then examined the sensitivity of the MOF rankings to the force field type. Our results showed that although there are significant quantitative differences between some adsorbent evaluation metrics computed using different force fields, rankings of the top MOF adsorbents for CO2 separations are generally similar: 8, 8, and 9 out of the top 10 most selective MOFs were found to be identical in the ranking for CO2/H2, CO2/N2, and CO2/CH4 separations using Dreiding and UFF. We finally suggested a force field factor depending on the energy parameters of atoms present in the MOFs to quantify the robustness of the simulation results to the force field selection. This easily computable factor will be highly useful to determine whether the results are sensitive to the force field type or not prior to performing computationally demanding molecular simulations.