Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS in Various Fields of Science and Engineering

CHAPTER 7

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS in Various Fields of Science and Engineering Sumit Sharma, Pramod Kumar, Rakesh Chandra Department of Mechanical Engineering, Dr. B.R. Ambedkar National Institute of Technology, Jalandhar, India

7.1 Applications of BIOVIA Materials Studio BIOVIA Materials Studio [1] is a complete modeling and simulation environment designed to allow researchers in materials science and chemistry to predict and understand the relationships of a material’s atomic and molecular structure with its properties and behavior. Using BIOVIA Materials Studio, researchers in many industries are engineering better-performing materials of all types, including pharmaceuticals, catalysts, polymers and composites, metals and alloys, batteries, and fuel cells. BIOVIA Materials Studio is the world’s most advanced yet easy-to-use environment for modeling and evaluating material performance and behavior. Using BIOVIA Materials Studio, materials scientists experience the following benefits: (a) Reduction in cost and time associated with physical testing and experimentation through “virtual screening” of candidate material variations. (b) Acceleration of the innovation process-developing new, better-performing, more sustainable, and cost-effective materials faster than can be done with physical testing and experimentation. (c) Improved fundamental understanding of the relationship between atomic and molecular structure with material properties and behavior. (d) Powerful material informatics capabilities through adoption of computational materials science as a complement to laboratory experimentation. (e) Automation and best-practice sharing with the BIOVIA Materials Studio Collection for BIOVIA Pipeline Pilot and the MaterialsScript API. Molecular Dynamics Simulation of Nanocomposites using BIOVIA Materials Studio, Lammps and Gromacs https://doi.org/10.1016/B978-0-12-816954-4.00007-3 # 2019 Elsevier Inc. All rights reserved.

329

330 Chapter 7 Chemists, polymer scientists, and other materials scientists become productive faster and with less effort using Materials Visualizer, the easiest to use and most complete graphic user environment for material modeling and simulation. Materials Visualizer provides capability to construct, manipulate, and view models of molecules, crystalline materials, surfaces, polymers, and mesoscale structures. It also supports the full range of BIOVIA Materials Studio simulations with capabilities to visualize results through images, animations, graphs, charts, tables, and textual data. Most tools in the Materials Visualizer can also be accessed through the MaterialsScript API, allowing expert users to create custom capabilities and automate repetitive tasks. Materials Visualizer’s Microsoft Windows client operates with a range of Windows and Linux server architectures to provide a highly responsive user experience. BIOVIA Materials Studio provides a complete range of simulation capabilities from quantum, atomistic, mesoscale, statistical, analytic, and crystallization tools. Its broad range of solutions enables researchers to evaluate materials at various particle sizes and timescales to predict properties more accurately and evaluate performance in the shortest time possible.

7.1.1 Quantum Tools (a) BIOVIA Materials Studio Cantera: Cantera (www.cantera.org) is a solver for chemical rate equations. BIOVIA Materials Studio Cantera provides environment for configuring the thermodynamic input and for executing these calculations. Cantera Reaction Editor enables users to introduce new species and reactions, optionally with reaction rates determined from BIOVIA Materials Studio DMol3 to complex reaction schemes with existing experimentally determined thermodynamic data. (b) BIOVIA Materials Studio CAmbridge Serial Total Energy Package (CASTEP): BIOVIA Materials Studio CASTEP simulates the properties of solids, interfaces, and surfaces for a wide range of materials including ceramics, semiconductors, and metals using a planewave density functional method. (c) BIOVIA Materials Studio DMol3: BIOVIA Materials Studio DMol3 is used to model the electronic structure and properties of organic and inorganic molecules, molecular crystals, covalent solids, metallic solids, and infinite surfaces using density functional theory (DFT). (d) BIOVIA Materials Studio Density Functional-based Tight Binding (DFTB+): BIOVIA Materials Studio DFTB+ is a semiempirical module for simulating electronic properties of materials. It uses a tight binding approach based on DFT to enable quantum mechanical accuracy on larger system sizes. (e) BIOVIA Materials Studio NMR CASTEP: BIOVIA Materials Studio NMR CASTEP predicts NMR chemical shifts and electric field gradient tensors from first principles. The method can be applied to compute the NMR shifts of both molecules and solids for a wide range of materials including ceramics and semiconductors.

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 331 (f ) BIOVIA Materials Studio Order-N Electronic Total Energy Package (ONETEP): BIOVIA Materials Studio ONETEP is a linear-scaling DFT code, enabling accurate, firstprinciple calculations on systems of up to thousands of atoms. (g) BIOVIA Materials Studio QMERA: BIOVIA Materials Studio QMERA employs quantum mechanics/molecular mechanics (QM/MM) method combining the accuracy of a quantum with the speed of a force field calculation. This approach makes it possible to perform accurate calculations on very large systems for substantially less effort. (h) BIOVIA Materials Studio VAMP: BIOVIA Materials Studio VAMP is capable of rapidly predicting many physical and chemical properties for molecular organic and inorganic systems using a semiempirical molecular orbital method. BIOVIA Materials Studio VAMP is an ideal intermediate approach between force field and first-principle methods.

7.1.2 Classical Simulation Tools BIOVIA Materials Studio offers a very wide range of methods based on classical interactions between atoms and molecules. These include molecular dynamics, lattice dynamics, and various Monte Carlo-based methods, as well as the provision of force fields: (a) BIOVIA Materials Studio Adsorption Locator: BIOVIA Materials Studio Adsorption Locator finds low-energy adsorption sites for molecules on both periodic and nonperiodic substrates. (b) BIOVIA Materials Studio Amorphous Cell: BIOVIA Materials Studio Amorphous Cell is a suite of computational tools that allow you to construct representative models of complex amorphous systems and to predict key properties. (c) BIOVIA Materials Studio Blends: BIOVIA Materials Studio Blends predicts phase diagrams and interaction parameters for liquid-liquid, polymer-polymer, and polymeradditive mixtures; phase equilibria; and separation technology. (d) BIOVIA Materials Studio Conformers: BIOVIA Materials Studio Conformers provides conformational search algorithms and analysis tools to characterize molecular conformation and flexibility. (e) BIOVIA Materials Studio Condensed-phase Optimized Molecular Potentials for Atomistic Simulation Studies (COMPASS): BIOVIA Materials Studio COMPASS is a force field that enables accurate prediction of structural, conformational, vibrational, and thermophysical properties for a broad range of molecules in isolation and in condensed phases and under a wide range of conditions of temperature and pressure. (f ) BIOVIA Materials Studio Forcite Plus: Forcite Plus offers molecular mechanics and dynamics methods for molecules and periodic systems. The tool includes a wide range of analysis features to predict mechanical properties, diffusivity, local structure, density variations, cohesive energy density, dipole autocorrelation functional, and more. Supported force fields are BIOVIA Materials Studio COMPASS, consistent valence force field (CVFF), polymer consistent force field (PCFF), Dreiding, and Universal.

332 Chapter 7 (g) BIOVIA Materials Studio GULP: GULP is a method for optimization, property calculation, and dynamics of materials. It includes a wide range of force fields for metals, oxides, minerals, semiconductors, and molecular mechanics force fields for covalent systems. Force field fitting tools are also provided to develop parameters for custom materials. (h) BIOVIA Materials Studio Sorption: Sorption provides a means of predicting fundamental properties needed for investigating adsorption and separations phenomena, such as sorption isotherms and Henry’s constants.

7.1.3 Mesoscale Simulation Tools Mesoscale methods in BIOVIA Materials Studio are based on a coarse-graining approach, whereby groups of atoms are replaced by beads. These methods enable the modeling of behavior at length and timescales that are beyond the range of classical tools: (a) MesoDyn: MesoDyn is a classical density functional method for studying the long-lengthscale and timescale behavior of complex fluid systems, in particular the phase separation and structure of complex polymer systems. (b) BIOVIA Materials Studio Mesocite: Mesocite is a coarse-grained simulation module for the study of materials at length scales ranging from nanometers to micrometers and timescales from nanoseconds to microseconds. BIOVIA Materials Studio Mesocite can provide structural and dynamic properties of fluids in equilibrium, under shear or in confined geometries.

7.1.4 Statistical Tools Statistical tools are ideal to screen compounds quickly by relating molecular traits directly to experimentally observed quantities: (a) BIOVIA Materials Studio quantitative structure-activity relationships (QSAR): QSAR’s integration in BIOVIA Materials Studio provides access to a wide range of descriptors and advanced analysis capabilities to help generate high-quality structure-activity relationships. QSAR includes a wide range of descriptors including topological and electrotopological descriptors. Also, Jurs descriptors enable charge distribution on solvent surfaces to be examined, VAMP descriptors further extend the range of 3-D descriptors into those including electronic interactions, and GFA applies a sophisticated genetic algorithm method to calculate quantitative structure-activity relationships. (b) BIOVIA Materials Studio QSAR Plus: QSAR Plus adds the power of the DMol3 descriptors for calculating reactivity indexes and accurate energies to QSAR. Also included are neural networks to build nonlinear models and models that are more resistant to noisy datasets than other model building methods. It can also be used with datasets that have some missing values and can be used to build weighted models to predict multiple physical properties.

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 333 (c) BIOVIA Materials Studio Synthia: Synthia calculates properties of homo- and copolymers using advanced quantitative structure-property relationships (QSPRs). It allows researchers to rapidly screen candidate polymers for a wide range of properties.

7.1.5 Analytical and Crystallization Tools Analytic and crystallization tools are employed to investigate, predict, and modify crystal structure and crystal growth: (a) BIOVIA Materials Studio Morphology: Morphology allows you to predict crystal morphology from the atomic structure of a crystal. Morphology allows for the prediction of crystal shape, the analysis of crystal surface stability, the development of tailor-made additives, and the control of solvent and impurity effects. (b) BIOVIA Materials Studio QSAR Plus: QSAR Plus adds the power of the DMol3 descriptors for calculating reactivity indexes and accurate energies to QSAR. Also included are neural networks to build nonlinear models and models that are more resistant to noisy datasets than other model building methods. It can also be used with datasets that have some missing values and can be used to build weighted models to predict multiple physical properties. (c) BIOVIA Materials Studio Motif: Motif analyzes connectivity information in molecular crystals, providing a qualitative and quantitative analysis method of hydrogen bond topologies. Combined with the predictive capabilities of Polymorph, Motif allows for categorization and statistical scoring of proposed structures. It interfaces with the Cambridge Structural Database exploiting Cambridge Crystallographic Data Centre’s Mercury functionality. (d) BIOVIA Materials Studio Reflex: Reflex simulates X-ray, neutron, and electron powder diffraction patterns based on models of crystalline materials. Reflex Plus offers a complete package for the determination of crystal structures from medium- to high-quality powder diffraction data.

7.2 Applications of LAMMPS Sorts of problems can be addressed using atomistic systems in LAMMPS [2] such as the following: (a) (b) (c) (d) (e)

Water interaction with self-assembled monolayers Ionomer morphologies Nanoparticle coating structures Self-assembly of lipid surfaces Soft material rheology

334 Chapter 7 (f ) (g) (h) (i) (j)

Wetting and surface properties of complex fluids Nanoindentation Corrosion behavior MD simulation of Li polymer batteries Prediction of mechanical, thermal, optical, and electronic properties of different types of composite materials

Some of the excellent studies discussed later highlight the importance of LAMMPS. Mackay et al. [3] implemented the long-range hydrodynamic interactions implemented into the open-source MD package, LAMMPS, through the creation of a fix, lb_fluid. These interactions were treated by interpolating the MD particle density onto a discrete lattice, which was then coupled to the fluid. A thermal lattice Boltzmann algorithm was used to model the fluid, which included mass- and momentum-conserving noise, providing a thermostat for both the particles and the fluid. Denniston et al. [4] implemented the Green’s function MD method, which enables one to study the elastic response of a three-dimensional solid to an external stress field by taking into consideration only the surface atoms, as an extension to an open-source classical MD simulation code LAMMPS. This was done in the style of fixes. The first fix, FixGFC, measured the elastic stiffness coefficients for a (small) solid block of a given material by making use of the fluctuation-dissipation theorem. With the help of the second fix, FixGFMD, the coefficients obtained from FixGFC were then used to compute the elastic forces for a (large) block of the same material. Both fixes were designed to be run in parallel and to exploit the functions provided by LAMMPS. Luo and Sommer [5] presented a patch code for LAMMPS to implement a CG model of polyvinyl alcohol (PVA). LAMMPS is a powerful MD simulator developed at Sandia National Laboratories. The patch code implements tabulated angular potential and Lennard-Jones-9-6 (LJ96) style interaction for PVA. Benefited from the excellent parallel efficiency of LAMMPS, this patch code is suitable for large-scale simulations. This CG-PVA code was used to study polymer crystallization, which was a long-standing unsolved problem in polymer physics. By using parallel computing, cooling and heating processes for long chains were simulated. The results showed that chain-folded structures resembling the lamellae of polymer crystals were formed during the cooling process. The evolution of the static structure factor during the crystallization transition indicated that long-range density order appeared before local crystalline packing. This was consistent with some experimental observations by small-/wide-angle X-ray scattering (SAXS/WAXS). During the heating process, it was found that the crystalline regions were still growing until they were fully melted, which could be confirmed by the evolution both of the static structure factor and average stem length formed by the chains. This two-stage behavior indicated that melting of polymer crystals was far from

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 335 thermodynamic equilibrium. The results concurred with various experiments. It was the first time that such growth/reorganization behavior was clearly observed by MD simulations. The code could be easily used to model other types of polymers by providing a file containing the tabulated angle potential data and a set of appropriate parameters. Urbassek et al. [6] presented a model for micrometer-sized grain-grain interaction, which exhibited the essential features necessary to describe collision, agglomeration, and fragmentation processes. The model was efficiently implemented in the LAMMPS code. In addition to existing models, adhesive forces and—gliding, rolling, and torsional—friction processes were implemented. The code scaled linearly with grain number up to 106 on a single processor and inherited the excellent parallelization features of the LAMMPS code on multiprocessor machines. The code was validated by determining the velocity of sound of granular agglomerates. Svaneborg [7] extended the LAMMPS to support directional bonds and dynamic bonding. The framework supported stochastic formation of new bonds, breakage of existing bonds, and conversion between bond types. Bond formation could be controlled to limit the maximal functionality of a bead with respect to various bond types. Concomitant with the bond dynamics, angular and dihedral interactions were dynamically introduced between newly connected triplets and quartets of beads, where the interaction type was determined from the local pattern of bead and bond types. When breaking bonds, all angular and dihedral interactions involving broken bonds were removed. The framework allowed chemical reactions to be modeled and use it to simulate a simplistic, CG DNA model. The resulting DNA dynamics illustrated the power of the present framework. A fully parallelized hybrid atomistic-continuum (HAC) model, built from the open-source codes LAMMPS and OpenFOAM, was developed by Lukes and Cosden [8] to resolve nanoscale features of fluid flow. The domain was decomposed into an atomistic domain, where individual atomic interactions were computed, and a continuum domain, where the Navier-Stokes equations were solved. The two domains were coupled through an overlap region in which the solutions in both domains were consistent. The accuracy of the HAC model was demonstrated through the simulation of sudden-start Couette flow. The hybrid model was shown to reduce computation time by a factor of five for a 78-nm channel as compared with a fully atomistic simulation; this speedup was expected to become even greater for larger systems. An accurate and efficient algorithm for calculating the 3-D pressure field was developed and implemented in the open-source MD package, LAMMPS, by Nakamura et al. [9]. Additionally an algorithm to compute the pressure profile along the radial direction in spherical coordinates was also implemented. The latter was particularly useful for systems showing a spherical symmetry such as micelles and vesicles. These methods yielded precise pressure fields based on the Irving-Kirkwood contour integration and were particularly useful for

336 Chapter 7 biomolecular force fields. The present methods were applied to several systems including a buckled membrane and a vesicle. Liu and Kuo [10] developed a driver module for executing volume-temperature replica exchange MD (VTREMD) for the LAMMPS package. As a patch code the VTREMD module performs classical MD with MC decisions between MD runs. The goal of inserting the MC step was to increase the breadth of sampled configurational space. In this method, states receive better sampling by making temperature or density swaps with their neighboring states. As an accelerated sampling method, VTREMD was particularly useful to explore states at low temperatures, where systems were easily trapped in local potential wells. As functional examples, TIP4P/Ew and TIP4P/2005 water models were analyzed using VTREMD. The phase diagram in this study covered the deeply supercooled regime, and this test served as a suitable demonstration of the usefulness of VTREMD in overcoming the slow dynamics problem. To facilitate using the current code, attention was also paid on how to optimize the exchange efficiency by using grid allocation. VTREMD was useful for studying systems with rough energy landscapes, such as those with numerous local minima or multiple characteristic timescales. Ma et al. [11] presented QMMMW, a new program aimed at performing QM/MM MD. The package operated as a wrapper that patches plane-wave self-consistent field (PWscf ) code included in the quantum ESPRESSO distribution and LAMMPS MD simulator. It was designed with a paradigm based on three guidelines: (i) minimal amount of modifications on the parent codes, (ii) flexibility and computational efficiency of the communication layer, and (iii) accuracy of the Hamiltonian describing the interaction between the QM and MM subsystems. These three features were seldom present simultaneously in other implementations of QMMM. The QMMMW project was hosted by qe-forge at (http://qe-forge. org/gf/project/qmmmw/). An analytic benchmark and a simple consistent Mathematica program were proposed by Favata et al. [12] for graphene and carbon nanotubes that might serve to test any MD code implemented with reactive empirical bond-order (REBO) potentials. By exploiting the benchmark the results produced by LAMMPS were checked when adopting the secondgeneration Brenner potential. It was made evident that this code in its current implementation produced results that were offset from those of the benchmark by a significant amount and provided evidence of the reason. Fu et al. [13] extensively studied the lipid bilayer membranes by CG MD simulations. Numerical efficiencies had been reported in the cases of aggressive coarse graining, where several lipids were CG into a particle of size 4–6 nm so that there was only one particle in the thickness direction. Yuan et al. [14] proposed a pair potential between these one-particle-thick CG lipid particles to capture the mechanical properties of a lipid bilayer membrane, such as gel-fluid-gas phase transitions of lipids, diffusion, and bending rigidity. Fu et al. [13] implemented such an interaction potential in LAMMPS to

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 337 simulate large-scale lipid systems such as a giant unilamellar vesicle (GUV) and red blood cells (RBCs). They also considered the effect of cytoskeleton on the lipid membrane dynamics as a model for RBC dynamics and incorporated CG water molecules to account for hydrodynamic interactions. The interaction between the CG water molecules (explicit solvent molecules) was modeled as a Lennard-Jones (L-J) potential. To demonstrate that the proposed methods did capture the observed dynamics of vesicles and RBCs, Fu et al. [13] focused on two sets of LAMMPS simulations: (i) vesicle shape transitions with enclosed volume and (ii) RBC shape transitions with different enclosed volume. Finally utilizing the parallel computing capability in LAMMPS, Fu et al. [13] provided some timing results for parallel CG simulations to illustrate that it was possible to use LAMMPS to simulate large-scale realistic complex biological membranes for more than 1 ms. The study of viscous fluid flow coupled with rigid or deformable solids has many applications in biological and engineering problems, for example, blood cell transport, drug delivery, and particulate flow. Tan et al. [15] developed a partitioned approach to solve this coupled multiphysics problem. The fluid motion was solved by parallel lattice Boltzmann solver (Palabos), while the solid displacement and deformation were simulated by LAMMPS. The coupling was achieved through the immersed boundary method (IBM). The code modeled both rigid and deformable solids exposed to flow. The code was validated with the Jeffery orbits of an ellipsoid particle in shear flow, RBC stretching test, and effective blood viscosity flowing in tubes. It demonstrated essentially linear scaling from 512 to 8192 cores for both strong and weak scaling cases. The computing time for the coupling increased with the solid fraction. An example of the fluid-solid coupling was given for flexible filament (drug carriers) transport in flowing blood cell suspensions, highlighting the advantages and capabilities of the developed code. Freitas et al. [16] described nonequilibrium techniques for the calculation of free energies of solids using MD simulations. These methods provided an alternative to standard equilibrium thermodynamic integration methods and often presented superior efficiency. The authors described the implementation in the LAMMPS code of two specific nonequilibrium processes that allowed the calculation of the free-energy difference between two different system Hamiltonians and the free-energy temperature dependence of a given Hamiltonian. The theory behind the methods was described (including fragments of LAMMPS scripts) and how the process parameters should be selected to obtain the best possible efficiency in the calculations of free energies using nonequilibrium MD simulations. As an example of the application of the methods the authors presented results related to polymorphic transitions for a classical potential model of iron. Leite and Koning [17] presented a guide to compute the absolute free energies of classical fluids using nonequilibrium free-energy techniques within the LAMMPS code. The main approach

338 Chapter 7 was based on the construction of a thermodynamic path connecting the fluid of interest to either atomic or molecular variants of the Uhlenbeck-Ford (UF) model as reference systems. These reference systems were described in detail, including their implementation in the LAMMPS package, and made available source code, scripts, and auxiliary files. As an illustration the authors detailed a number of distinct applications, involving system characterized by fundamentally different interactions. In addition to two different atomic models (model of water (mW) and the modified embedded atom method (MEAM)-second nearest neighbor (2NN) CuZr liquid binary alloy) the authors considered three molecular models for water, two of them rigid (TIP4P and SPC/E) and one flexible (q-SPC/Fw). For the molecular models the authors developed UF-based reference systems for which the free energies were given by a sum of two contributions: an intermolecular part described by the known UF free energy and an intramolecular contribution that could be determined analytically. The tools described in this study provided a platform on which fluid-phase free energies could be easily and efficiently computed using the LAMMPS code. In addition to being useful for the development of new models for liquid phases the tools may also find applications in the construction of community databases containing thermodynamic properties of existing models. From the earlier studies, it can be inferred that LAMMPS is in use for solving problems in almost every field of engineering whether it is medical, mechanical, electronic, chemical, or any other field. The readers are requested to go through the references given in this chapter for better understanding of LAMMPS.

7.3 Applications of GROMACS The following paragraphs will explain the utility of GROMACS in various areas. Padmanabhan et al. [18] explored an eco-friendly method for the production of silver nanoparticles from Bacillus clausii cultured from Enterogermina. Along with the biosynthesis and conformity test, in silico studies were done on nicotinamide adenine dinucleotide phosphate (NADPH)dependent nitrate reductase enzymes from the viewpoint of designing a rational enzymatic strategy for the synthesis. The detailed characterization of the nanoparticles was carried out using UV-vis spectroscopy, dynamic light scattering (DLS) particle size analysis, transmission electron microscopy (TEM), and x-ray diffraction (XRD) analysis. Computational profiling and in silico characterization of NADH-dependent enzymes were carried out based on literature. Nitrate reductase sequence was retrieved from National Center for Biotechnology Information (NCBI) for characterization. Secondary structure was evaluated and verified by JPred and Self-Optimized Prediction Method with Alignment (SOPMA) tool. Tertiary structure was also modeled by MODELLER and Iterative Threading ASSEmbly Refinement (I-TASSER) parallel, and the best structure was selected based on energy values.

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 339 Structure validation was done by GROMACS, and root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), and temperature variation plot were also plotted. Interaction graphs between nitrate reductase and ligand silver nitrate were done through molecular docking using Hex. Vattulainen et al. [19] provided a data package of GROMACS input files for atomistic MD simulations of multicomponent, asymmetrical lipid bilayers using the optimized potentials for liquid simulations-all atom (OPLS-AA) force field. The data included 14 model bilayers composed of eight different lipid molecules. The lipids present in these models were cholesterol (CHOL), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC), 1-palmitoyl-2oleoyl-sn-glycero-3-phosphatidylethanolamine (POPE), 1-stearoyl-2-oleoyl-snglycero-3-phosphatidylethanolamine (SOPE), 1-palmitoyl-2-oleoyl-snglycero-3-phosphatidylserine (POPS), 1-stearoyl-2-oleoyl-sn-glycero-3-phosphatidylserine (SOPS), N-palmitoyl-D-erythro-sphingosyl-phosphatidylcholine (SM16), and N-lignoceroy l-D-erythro-sphingosyl-phosphatidylcholine (SM24). The bilayers’ compositions were based on lipidomic studies of PC-3 prostate cancer cells and exosomes discussed in Llorente et al. [20], showing an increase in the section of long-tail lipid species (SOPS, SOPE, and SM24) in the exosomes. Former knowledge about lipid asymmetry in cell membranes was accounted for in the models, meaning that the model of the inner leaflet was composed of a mixture of PC, PS, PE, and cholesterol, while the extracellular leaflet was composed of SM, PC, and cholesterol discussed in Van Meer et al. [21]. The provided data include lipids’ topologies, equilibrated structures of asymmetrical bilayers, all force field parameters, and input files with parameters describing simulation conditions (md.mdp). The data were associated with the research article by Ro´g et al. [22]. Ro´g et al. [23] provided topologies and force field parameter files for MD simulations of lipids in the OPLS-AA force field using the GROMACS package. This was the first systematic parameterization of lipid molecules in this force field. Topologies were provided for four phosphatidylcholines: saturated dipalmitoylphosphatidylcholine (DPPC), mono-cisunsaturated POPC and 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), and mono-transunsaturated phosphoenolpyruvate carboxylase (PEPC). Parameterization of the phosphatidylcholines was achieved in two steps: first the authors supplemented the OPLS force field parameters for DPPC with new parameters for torsion angles and van der Waals parameters for the carbon and hydrogen atoms in the acyl chains and new partial atomic charges and parameters for torsion angles in the phosphatidylcholine and glycerol moieties [24]. Next, parameters were derived for the cis- and transdouble bonds and the neighboring single bonds [25]. Additionally the authors provided GROMACS input files with parameters describing simulation conditions (md.mdp), which were strongly recommended to be used with these lipid models. The data were associated with the research article by Kulig et al. [25] and provided as supporting materials.

340 Chapter 7 In the study by Lemkul [26], users are provided the rationale and a theoretical explanation for the command-line syntax in each step in the online tutorials (available at http://www. mdtutorials.com/gmx) and the underlying settings and algorithms necessary to perform robust MD simulations in each scenario.

References [1] Dassault Syste`mes BIOVIA, Materials Studio, 7.0, Dassault Syste`mes, San Diego, 2017. [2] https://icme.hpc.msstate.edu/mediawiki/index.php/LAMMPS_tutorials. [3] F.E. Mackay, S.T.T. Ollila, C. Denniston, Hydrodynamic forces implemented into LAMMPS through a latticeBoltzmann fluid, Comput. Phys. Commun. 184 (2013) 2021–2031. [4] C. Denniston, L.T. Kong, G. Bartels, C. Campan˜a´, M.H. M€ user, Implementation of Green’s function molecular dynamics: an extension to LAMMPS, Comput. Phys. Commun. 180 (2009) 1004–1010. [5] C. Luo, J.U. Sommer, Coding coarse grained polymer model for LAMMPS and its application to polymer crystallization, Comput. Phys. Commun. 180 (2009) 1382–1391. [6] H.M. Urbassek, C. Ringl, A LAMMPS implementation of granular mechanics: inclusion of adhesive and microscopic friction forces, Comput. Phys. Commun. 183 (2012) 986–992. [7] C. Svaneborg, LAMMPS framework for dynamic bonding and an application modeling DNA, Comput. Phys. Commun. 183 (2012) 1793–1802. [8] J.R. Lukes, I.A. Cosden, A hybrid atomistic-continuum model for fluid flow using LAMMPS and OpenFOAM, Comput. Phys. Commun. 184 (2013) 1958–1965. [9] T. Nakamura, S. Kawamoto, W. Shinoda, Precise calculation of the local pressure tensor in Cartesian and spherical coordinates in LAMMPS, Comput. Phys. Commun. 190 (2015) 120–128. [10] L.C. Liu, J.L. Kuo, A LAMMPS implementation of volume-temperature replica exchange molecular dynamics, Comput. Phys. Commun. 189 (2015) 119–127. [11] C. Ma, L. Martin-Samos, S. Fabris, A. Laio, S. Piccinin, QMMMW: a wrapper for QM/MM simulations with Quantum ESPRESSO and LAMMPS, Comput. Phys. Commun. 195 (2015) 191–198. [12] A. Favata, A. Micheletti, S. Ryu, N.M. Pugno, An analytical benchmark and a Mathematica program for MD codes: Testing LAMMPS on the 2nd generation Brenner potential, Comput. Phys. Commun. 207 (2016) 426–431. [13] S.P. Fu, Z. Peng, H. Yuan, R. Kfoury, Y.N. Young, Lennard-Jones type pair-potential method for coarsegrained lipid bilayer membrane simulations in LAMMPS, Comput. Phys. Commun. 210 (2017) 193–203. [14] H. Yuan, C. Huang, J. Li, G. Lykotrafitis, S. Zhang, One-particle-thick, solvent-free, coarse-grained model for biological and biomimetic fluid membranes, Phys. Rev. E 82 (2010) 011905. [15] J. Tan, T.R. Sinno, S.L. Diamond, A parallel fluid-solid coupling model using LAMMPS and Palabos based on the immersed boundary method, J. Comput. Sci. 25 (2018) 89–100. [16] R. Freitas, M. Asta, M. de Koning, Nonequilibrium free-energy calculation of solids using LAMMPS, Comput. Mater. Sci. 112 (2016) 333–341. [17] R.P. Leite, M. de Koning, Nonequilibrium free-energy calculations of fluids using LAMMPS, Comput. Mater. Sci. 159 (2019) 316–326. [18] P. Padmanabhan, K. Mukherjee, R. Gupta, G. Kumar, S. Kumari, S. Biswas, Synthesis of silver nanoparticles by Bacillus clausii and computational profiling of nitrate reductase enzyme involved in production, J. Genet. Eng. Biotechnol. 16 (2018) 527–536. [19] I. Vattulainen, T. Ro´g, A. Orłowski, et al., Data including GROMACS input files for atomistic molecular dynamics simulations of mixed, asymmetric bilayers including molecular topologies, equilibrated structures, and force field for lipids compatible with OPLS-AA parameters, Data Brief 7 (2016) 1171–1174. [20] A. Llorente, I. Vattulainen, T. Ro´g, et al., Molecular lipidomics of exosomes released by PC-3 prostate cancer cells, Biochim. Biophys. Acta 1831 (2013) 1302–1309.

Applications of BIOVIA Materials Studio, LAMMPS, and GROMACS 341 [21] G. van Meer, D.R. Voelker, G.W. Feigenson, Membranes lipids: where they are and how they behave, Nat. Rev. Mol. Cell Biol. 9 (2008) 112–124. [22] T. Ro´g, A. Orłowski, A. Llorente, et al., Interdigitation of long-chain sphingomyelin induces coupling of membrane leaflets in a cholesterol dependent manner, Biochim. Biophys. Acta 1858 (2016) 281–288. [23] T. Ro´g, W. Kulig, M.P. Gierula, Topologies, structures and parameter files for lipid simulations in GROMACS with the OPLS-aa force field: DPPC, POPC, DOPC, PEPC, and cholesterol, Data Brief 5 (2015) 333–336. [24] A. Maciejewski, M. Pasenkiewicz-Gierula, O. Cramariuc, I. Vattulainen, T. Ro´g, Refined OPLS-AA force field for saturated phosphatidylcholine bilayers at full hydration, J. Phys. Chem. B 118 (2014) 4571–4581. [25] W. Kulig, M. Pasenkiewicz-Gierula, T. Ro´g, Cis and trans unsaturated phosphatidylcholine bilayers: a molecular dynamics simulation study, Chem. Phys. Lipids 195 (2016) 12–20. [26] J.A. Lemkul, From proteins to perturbed Hamiltonians: a suite of tutorials for the GROMACS-2018 molecular simulation package, Living J. Comput. Mol. Sci. 1 (1) (2019) 5068.