State of the Art in Mesoscale and Multiscale Modeling

State of the art in mesoscale and multiscale modeling

May 29, 2017 - June 01, 2017 CECAM-IRL

Ignacio Pagonabarraga CECAM EPFL, Switzerland

Vladimir Lobaskin University College Dublin, Ireland

Donal Mac Kernan University College Dublin, Ireland

https://www.cecam.org/workshop-8-1487.html Program

Day 1 – 29 May, 2017 UCD School of Physics Science North Room 232 - shuttle bus (departs hotel at 9.30 for UCD) max 16 passangers google map link https://goo.gl/maps/LQ9pTNMtYKv •10:00 to 11:00 - Didactic Lecture I methodologies of systematic static coarse- graining •11:00 to 11:30 - Coffee Break •11:30 to 12:30 - Burkhard Duenweg Introduction to the Mori-Zwanzig formalism •12:30 to 13:30 - Lunch •13:30 to 14:10 - Overview Talk on mesoscale/hybrid approaches (KREMER) •14:10 to 14:30 - Christian Holm Influence of the permittivity gradient on static and dynamic properties of charged macromolecules •14:30 to 14:50 - Burkhard Duenweg Monte Carlo approach to Fluctuating Lattice Boltzmann •14:50 to 15:10 - Matej Praprotnik Adaptive resolution simulations of supramolecular water •15:10 to 15:30 - Pietro Lio Multiscale computational approaches in modeling biological data integration and molecular communication •15:30 to 16:00 - Coffee Break •16:00 to 16:20 - Pierre Cazade Multi-scale modelling of Large Biomolecular Complexes •16:20 to 16:40 - Focused/Question specific talk 5 on mesoscale/hybrid approaches •16:40 to 17:00 - Jason Reese Best of Both Worlds? Hybrid Fluid Simulations for Multiscale Engineering •17:00 to 17:20 - Gerhard Jung Iterative Reconstruction of Memory Kernels •17:20 to 18:20 - Roundtable on current challenges for hybrid approaches, inclusion of ab- initio description (moderator Kremer) •18:20 to 19:50 - Reception-welcome-poster session (Physics Lounge) Day 2 – 30 May, 2017 UCD School of Physics Science North Room 232 shuttle bus departs hotel at 9.0 & 9.15 for UCD max 16 passangers •09:30 to 10:10 - Overview on non-equilibrium processes and heterogeneous systems (RIEGER) •10:10 to 10:30 - Julija Zavadlav Multiscale Simulations of DNA Arrays •10:30 to 10:50 - Aleksandar Donev Coupling a nano-particle with fluctuating hydrodynamics •10:50 to 11:20 - Coffee Break •11:20 to 11:40 - David Cheung Coarse-grain modeling of polymer nanostructures •11:40 to 12:30 - Roundtable on non-equilibrium processes (moderator Reiger) •12:30 to 13:30 - Lunch •13:30 to 14:10 - Overview on Coarse Graining (MUELLER) •14:10 to 14:30 - Alexander Lyubartsev Multiscale modelling by systematic structure-based coarse-graining •14:30 to 14:50 - Agur Sevink Automated multi-scaling with Stochastic Quasi-Newton (S-QN) •14:50 to 15:10 - Pep Español Non-isothermal coarse-graining of complex molecules •15:10 to 15:30 - Denis Andrienko Many-body effects in systematic coarse graining •15:30 to 15:50 - Nicolae-Viorel Buchete Coarse-Grained Kinetic Models of Protein Folding and Binding Networks •15:50 to 16:20 - Coffee Break •16:20 to 17:00 - Overview on connecting to continuum/engineering level descriptions (Casciola) •17:00 to 17:20 - Pietro Asinari Multi-Scale Modelling of Nanoparticle Suspensions •17:20 to 17:40 - Serafim Kalliadasis From the nano- to the macroscale: bridging scales for the moving contact line problem •17:40 to 18:40 - Roundtable on connecting to continuum/engineering level descriptions (moderator O'REILLY) •19:45 to 22:00 - Workshop Dinner on Canal Boat (lapeniche.ie) Day 3 – 31 May, 2017 UCD School of Physics Science North Room 232 shuttle bus departs hotel at 9.0 & 9.15 for UCD max 16 passangers •09:30 to 10:10 - Overview on Software packages (SEATON) •10:10 to 10:30 - Kevin Stratford Performance portability in coarse-grained mesoscale complex fluids •10:30 to 10:50 - Danny Perez Long-timescale Simulations with Accelerated Molecular Dynamics and the EXAALT package •10:50 to 11:20 - Coffee Break •Moderated roundtable on software solutions and their limitations: H2020 funding opportunities (ASINARI) •12:30 to 13:30 - Lunch •13:30 to 14:10 - Additional Research talks/AOB •14:10 to 15:10 - Denis Andrienko Systematic coarse-graining with VOTCA •15:10 to 15:40 - Coffee Break •15:40 to 16:40 - Didactic Lecture on continuum/engineering level descriptions 3 Abstracts

Introduction to the Mori-Zwanzig formalism Burkhard Duenweg[1] Max Planck Institute for Polymer Research Mainz

The lecture will outline the basic steps to derive the Mori-Zwanzig memory equation, and illustrate its application to derive some well-known Green-Kubo relations. If time permits, the method of multiple time scale expansion (aka Chapman-Enskog expansion within the framework of the kinetic theory of gases) will be outlined as well. Both approaches have in common that they aim at deriving effective equations of motion for slow degrees of freedom, which often involves to study the system not only on large time scales, but on large length scales as well.

Influence of the permittivity gradient on static and dynamic properties of charged macromolecules Christian Holm[1] Institut für Computerphysik, Universität Stuttgart, Stuttgart, Germany

Dissolved ions can alter the local permittivity of water, nevertheless most theories and simulations ignore this fact. We present a novel algorithm for treating spatial and temporal variations in the permittivity, and show several examples where this leads to large qualitative and quantitative differences. A dynamic example is the equivalent conductivity of a salt-free polyelectrolyte solution. Our new approach quantitatively reproduces experimental results unlike simulations with a constant permittivity that even qualitatively fail to describe the data. We can relate this success to a change in the ion distribution close to the polymer due to the built-up of a permittivity gradient.

F. Fahrenberger, O.A. Hickey, J. Smiatek, C. Holm, „The influence of charged-induced variations in the local permittivity on the static and dynamic properties of polyelectrolyte solutions“. J. Chem. Phys. 143, 243140, (2015). F. Fahrenberger, O.A. Hickey, J. Smiatek, C. Holm, „Importance of Varying Permittivity on the Conductivity of Polyelectrolyte Solutions“, Phys. Rev. Lett.115, 118301 (2015). F. Fahrenberger, C. Holm, „Computing the Coulomb interaction in inhomogeneous dielectric media via a local electrostatics lattice algorithm“ Phys. Rev. E 90, 063304, (2014). F. Fahrenberger, Z. Xu, C. Holm, „Simulation of electric double layers around charged colloids in aqueous solution of variable permittivity“, J. Chem. Phys. 141 064902 (2014)

Monte Carlo approach to Fluctuating Lattice Boltzmann Burkhard Duenweg[1], Thomas Balles, Benjamin Pampel Max Planck Institute for Polymer Research Mainz TU Darmstadt

We explore possibilities to sample the populations of a Fluctuating Lattice Boltzmann algorithm according to the correct Boltzmann distribution instead of the usually-applied Gaussian approximation thereof. Since there is no straightforward method to generate the needed random variables directly, we use Markov Chain Monte Carlo methods instead.

Adaptive resolution simulations of supramolecular water Matej Praprotnik[1] National Institute of Chemistry, Ljubljana, Slovenia

We present adaptive resolution molecular dynamics simulations of water using coarse-grained molecular models that are compatible with the MARTINI force field [1,2]. The solvent molecules change their resolution back and forth between the atomistic and coarse-grained representations according to their positions in the system. The difficulties that arise from coupling to a coarse-grained model with a supramolecular mapping are successfully circumvented by using bundled water models. We discuss the advantages and limitations of this multiscale approach on several examples, e.g., coupling of atomistic water with polarizable [3] and non-polarizable [4] coarse-grained water models. To overcome the limitations of the bundled water models we then intruduce a dynamic clustering algorithm SWINGER [5] that can concurrently assemble, disassemble, and reassemble water bundles, consisting of several water molecules. It allows for a seamless coupling between standard atomistic and supramolecular water models in adaptive resolution simulations. SWINGER paves the way for efficient multiscale simulations of biomolecular systems without compromising the accuracy of atomistic water models.

[1] J. Zavadlav, M. N. Melo, S. J. Marrink, M. Praprotnik, J. Chem. Phys. 140, 054114, 2014. [2] J. Zavadlav, R. Podgornik, M. N. Melo, S. J. Marrink, M. Praprotnik, Eur. Phys. J. Special Topics 225, 1595-1607, 2016. [3] J. Zavadlav, M. N. Melo, S. J. Marrink, M. Praprotnik, J. Chem. Phys. 142, 244118, 2015. [4] J. Zavadlav, M. N. Melo, A. V. Cunha, A. H. de Vries, S. J. Marrink, M. Praprotnik, J. Chem. Theory Comput. 10, 2591-2598, 2014. [5] J. Zavadlav, S. J. Marrink, M. Praprotnik, J. Chem. Theory Comput. 12, 4138-4145, 2016.

Multiscale computational approaches in modeling biological data integration and molecular communication Pietro Lio[1] University of Cambridge

I have developed multiscale data integration computational approaches based on hierarchical block matrices [1], multilayer networks [2,3], cross-modal convolution networks [4], matrix factorisation and tri-factorisation. These methodologies (implemented in software tools used by a large community) combine clinical data from longitudinal studies and multidimensional molecular data (genomics, metabolomics, proteomics, metagenomics and epigenomics) with systems biomedicine approaches (network modeling, survival analysis, immune system simulations) to identify meaningful sub-categories of patients and their therapy responsiveness. [5]

[1] Y. Shavit, B. Walker, P. Lio' Bioinformatics 2015 Dec 17. pii: btv736 , 2015. [2] C. Angione, M. Conway and P Lio' BMC Bioinformatics. 17 Suppl 4:83. doi: 10.1186/s12859-016-0912- 1, 2016. [3] P. Veličković, P. Lio', Journal of Complex Networks; doi: 10.1093/comnet/cnv029, 2015. [4] P. Veličković, D. Wang, N. Lane, P. Lio' IEEE SSCI 2016: 1-8, 2016 [5] G. Ascolani, A. Occhipinti, P. Lio’ Plos Computational Biology May 15;11(5):e1004199. doi: 10.1371/journal.pcbi.1004199, 2015.

Multi-scale modelling of Large Biomolecular Complexes Pierre Cazade[1], Damien Thompson University of Limerick University of Limerick

We are using quantum mechanically-derived molecular dynamics force fields together with coarse grained models and free energy simulations to guide the development of so-called "designer cellulosome" bio-catalysts capable of fast and cost-effective breakdown of feedstock waste for biofuel production. These complex biological nanomachines are composed of multiple proteins connected by flexible linker regions. Such a complex design problem not only challenges supercomputing capabilities but also exploits the latest developments in out-of- equilibrium simulation methods, accelerated sampling, and analysis methodologies to extract relevant information from very large datasets. Our data shows that naturally-occurring glycosylation serves to keep the protein units in their active forms and at optimum distances from each other. We show that the units can be rationally re-engineered into states of lower and higher mechanostability, as very recently validated by our experimental collaborators using force microscopy and spectroscopy measurements. Best of Both Worlds? Hybrid Fluid Simulations for Multiscale Engineering Jason Reese[1], Matthew Borg University of Edinburgh, UK University of Edinburgh, UK

Micro- and nano-scale fluid systems can behave very differently from their macro-scale counterparts. Molecular effects require special treatment beyond the scope of conventional continuum-fluid modelling. In this talk I will describe the hybrid molecular/continuum schemes we are constructing to tackle these multiscale problems. I will discuss domain-decomposition vs heterogeneous techniques, and sequential vs concurrent methods. The challenge is to balance improved accuracy with computational practicality. To illustrate the different approaches, I will present examples of engineering applications, including the materials design of nanotube filtration membranes.

Iterative Reconstruction of Memory Kernels Gerhard Jung[1], Friederike Schmid University of Mainz, Germany University of Mainz, Germany

In recent years it has become increasingly popular to construct coarse-grained models with Non-Markovian dynamics in order to account for an incomplete separation of time scales. One challenge of a systematic coarse-graining procedure is the extraction of the dynamical properties, namely the memory kernel, from all-atom simulations. In this talk we will present the iterative memory reconstruction (IMR) [1], which is the first iterative method suggested in the field of Non-Markovian modeling. This new extraction technique shows various similarities to the reconstruction of static potentials via inverse Monte Carlo (IMC) [2] or iterative Boltzmann inversion (IBI) [3] . The big advantage compared to previously proposed techniques (e.g. [4,5]) is the efficient reconstruction of coarse-grained models that have precisely the same dynamical properties as the original fine-grained systems. We demonstrate the performance of the above described methods using the example of backflow induced memory in the Brownian diffusion of a single colloid. For this system we are able to reconstruct realistic coarse-grained dynamics with time steps about 200 times larger than used in the original molecular dynamics simulations.

[1] G. Jung and F. Schmid, submitted to JCTC [2] A. P. Lyubartsev and A. Laaksonen, Phys. Rev. E 52, 3730 (1995) [3] D. Reith, M. Pütz, F. Müller-Plathe, Comp. Chem. 24, 1624 (2003) [4] H. K. Shin, C. Kim, P. Talkner and E. K. Lee, Chem. Phys. 375, 316 (2010) [5] A. Carof, R. Vuilleumier and B. Rotenberg, J. Chem. Phys. 140, 124103 (2014)

Multiscale Simulations of DNA Arrays Julija Zavadlav[1], Rudolf Podgornik [1, 2], Matej Praprotnik [2, 3] ETH Zurich, Zurich, Switzerland [1] J. Stefan Institute, Ljubljana, Slovenia [2] University of Ljubljana, Ljubljana, Slovenia [3] National Institute of Chemistry, Ljubljana, Slovenia

We present recent biomolecular applications of the adaptive resolution scheme (AdResS), which allows solvent molecules to change their resolution back and forth between atomistic and coarse-grained representations according to their positions in the system. First, we discuss coupling of atomistic and coarse-grained models of salt solution, which is later applied in a multiscale simulation of a DNA molecule in 1 M NaCl salt solution environment [1]. The region of high resolution moves together with the DNA center-of-mass so that the DNA itself is always modeled at high resolution. We show that our multiscale simulations yield a stable DNA-solution system, with statistical properties similar to those produced by the conventional all-atom molecular dynamics simulation. In the second application, we perform a multiscale simulation of dense DNA arrays by enclosing a set of atomistically resolved DNA molecules within a semi- permeable membrane [2], allowing the passage of water and salt ions, and thus mimicking an open system setup, using an open boundary molecular dynamics (OBMD) -like approach [3,4]. By varying the DNA density, local packing symmetry, and counterion type, we obtain the osmotic equation of state together with the hexagonal-orthorhombic phase transition. Our multiscale approach opens the way for large-scale applications of DNA and other biomolecules, which require a large solvent reservoir. [1] J. Zavadlav, R. Podgornik, and M. Praprotnik, J. Chem. Theory Comput. 11 5035 (2015) [2] J. Zavadlav, R. Podgornik, and M. Praprotnik, submitted [3] R. Delgado-Buscalioni, J. Sablić and M. Praprotnik, Eur. Phys. J. Spec. Top. 224 2331 (2015) [4] J. Sablić, M. Praprotnik, and R. Delgado-Buscalioni, Soft Matter 12 2416-2439 (2016)

Coupling a nano-particle with fluctuating hydrodynamics Aleksandar Donev[1], Pep Espanol Courant Institute, NYU UNED, Madrid, Spain

We derive a coarse-grained description of the dynamics of a nanoparticle immersed in an isothermal simple fluid by performing a systematic coarse graining of the underlying microscopic dynamics. As coarse-grained or relevant variables we select the position of the nanoparticle and the total mass and momentum density field of the fluid, which are locally conserved slow variables because they are defined to include the contribution of the nanoparticle. The theory of coarse graining based on the Zwanzing projection operator leads us to a system of stochastic ordinary differential equations (SODEs) that are closed in the relevant variables. We demonstrate that our discrete coarse-grained equations are consistent with a Petrov-Galerkin finite-element discretization of a system of formal stochastic partial differential equations (SPDEs) which resemble previously-used phenomenological models based on fluctuating hydrodynamics. Under suitable approximations we obtain closed approximations of the coarse- grained dynamics in a manner which gives them a clear physical interpretation, and provides explicit microscopic expressions for all of the coefficients appearing in the closure. Our work leads to a model for dilute nanocolloidal suspensions that can be simulated effectively using feasibly short molecular dynamics simulations as input to a FEM fluctuating hydrodynamic solver.

"Coupling a nano-particle with isothermal fluctuating hydrodynamics: Coarse-graining from microscopic to mesoscopic dynamics" P. Español and A. Donev, J. Chem. Phys., 143, 234104, 2015 [ArXiv:1509.01540].

Coarse-grain modeling of polymer nanostructures David Cheung[1] School of Chemistry, National University of Ireland Galway

Polymer vesicles, fluid-filled polymer sacs, have attracted great interest for a range of applications, such as drug delivery vehicles or miniature chemical reactors. Due to their similarity to biological cells they are often considered as minimal synthetic models of cells. Compared to naturally occurring systems synthetic vesicles are significantly simpler, lacking the much of the complexity of biological cells, which arises due to, e.g. the heterogeneous composition of the cell membrane or the attachment of proteins. Recently a number of experimental studies have begun to incorporate some of this complexity seen in biological systems, through the creation of multicomponent vesicles or vesicles armoured with colloidal particles. However, these experimental studies have not characterized in detail the microscopic changes in bilayer structure and properties.This presentation will discuss two pieces of simulation work that try to address some of this missing information. In the first simple Monte Carlo simulations were used to study the packing of colloidal particles onto polymer vesicles [1,2]. These simulations were able to reproduce the packing patterns seen in these experimental systems and to study the effect of surface charge density on the packing pattern. The second will address the effect of nanoparticles on the structure and phase separation in mixed polymer bilayers [2]. Through establishing the relationship between the properties of the nanoparticles and the structure of polymer vesicles these simulation results will give insight into the directed assembly of polymer-nanoparticle structures, as well as giving insight into biological membranes.

[1] R Chen, DJG Pearce, S Fortuna, DL Cheung, and SAF Bon, J. Am. Chem. Soc., 133, 2151-2153 (2011) [2] S Fortuna and DL Cheung, Coll. Inter. Sci. Comm., 17, 10-13 (2017) [3] DL Cheung, J. Chem. Phys., 141, 194908 (2014) Multiscale modelling by systematic structure-based coarse-graining Alexander Lyubartsev[1] UNiversity Stockholm

TBA

Automated multi-scaling with Stochastic Quasi-Newton (S-QN) Agur Sevink[1] University of Leiden

When averaging over less significant degrees of freedom and mapping from a fine to a coarser molecular representation, one important challenge associated with statics is that the original potential energy (hyper)surface or PES needs to be consistently mapped onto an equivalent (and smoother) PES of the coarser description. An attractive alternative for such a mapping is to exploit the topographic features of the original PES to increase the sampling rate, enabling a much more extensive search for states that are significant for the equilibrium behaviour. In this talk, I present a recently developed S-QN method that, regardless of the molecular representation, enables just that by efficiently building Hessian information of the PES along the sampling trajectory. This Hessian information can either be instantly used to enhance the sampling rate, via an automated assignment of larger step sizes for cooperative modes, or employed as a post-processing routine to extract cooperativity. The latter gives access to information needed for the proper development of coarser representations. As a method for enhanced sampling, S-QN is an attractive alternative for non-directional Monte Carlo and also for the familiar metadynamics, which takes a different route to enable enhanced sampling, but has the disadvantage that the cooperative degrees of freedom should be predefined.

Non-isothermal coarse-graining of complex molecules Pep Español[1] UNED

Coarse-graining (CG) of complex molecules is a way to reach timescales that would be impossible to access through brute forcemolecular simulations. We formulate a coarse- grained model forcomplex molecules from first principles that ensures energyconservation. Each molecule is described in a coarse way by a thermalblob characterized by the position and momentum of the center of massof the molecule, together with its internal energy as an additionaldegree of freedom. This level of description gives rise to anentropy-based framework instead of the usual one based on theconfigurational free energy (i.e. potential of mean force). Theresulting dynamic equations, which account for a proper description ofheat transfer at the coarse grained level, have the structure of theDissipative Particle Dynamics with Energy conservation (DPDE) modelbut with a clear microscopic underpinning. Under suitableapproximations, we provide explicit microscopic expressions for eachcomponent (entropy, mean force, friction and conductivitycoefficients) appearing in the coarse-grained model. These quantitiescan be computed directly from MD simulations. We show, in particular, that it is possible to talk about "the entropy of a single molecule". The proposednon-isothermal coarse-grained model is thermodynamically consistentand opens up a first principles CG strategy to the study of energytransport issues that are not accessible with current isothermalmodels. Many-body effects in systematic coarse graining Denis Andrienko[1], Christoph Scherer Max Planck Institute for Polymer Research

Particle based coarse-graining (CG) is a systematic way of reduction of the number of degrees of freedom describing a physical system. It involves three steps: choice of the CG degrees of freedom, identification of a merit function which quantifies the difference between the fine- and coarse-grained representations, and determination of the CG potential energy surface (PES). The entire procedure is sensitive to the number and types of basis-functions employed in the CG representations: for example, the incorporation of nonbonded three-body interactions in a coarse-grained water model helps to reproduce both thermodynamic and structural properties [1,2]. In this work, we investigate the effect of extending the basis set to three-body interactions for several organic solvents and formulate clear criteria for when these many-body terms are required. The coarse-graining scheme is implemented in the VOTCA-CSG toolkit [3].

[1] V. Molinero, E. B. Moore, J. Phys. Chem. B, 113, 4008 (2009) [2] L. Larini, L. Lu, G.A. Voth, J. Chem. Phys. 132, 164107 (2010) [3] V. Rühle, C. Junghans, A. Lukyanov, K. Kremer, D. Andrienko, J. Chem. Theory. Comput. 5, 3211 (2009)

Coarse-Grained Kinetic Models of Protein Folding and Binding Networks Nicolae-Viorel Buchete[1] University College Dublin, Ireland

Protein folding and binding processes involve intrinsically complex mechanisms and corresponding pathways, which are notoriously difficult to study exhaustively, in a systematic manner. We show how master equations can be constructed and used to extract accurate coarse-grained kinetic models of the conformational dynamics of biomolecules, based on data from atomistic replica-exchange molecular dynamics simulations with explicit water molecules. By carefully controlling the effects of fast but often non-Markovian transitions, on one hand, and the typically limited sampling of slow relaxation processes on the other hand, we probe the underlying network of folding-unfolding and binding-unbinding transitions between the various configuration states of our biomolecular system. This systematic analysis reveals the important intrinsic conformational states and the associated transition pathways at multiple levels, from atomistic to coarse-grained representations. We validate our approach in several MD studies of amyloid-forming peptides, and of larger systems such as Cyclosporin A, and the helix-turn-helix subdomain of a viral scaffolding protein.

[1] C.T. Leahy, R.D. Murphy, G. Hummer, E. Rosta, and N.V. Buchete, J. Phys. Chem. Lett. 7, 2676 (2016). [2] L. Martini, A. Kells, G. Hummer, N.V. Buchete, and E. Rosta, arXiv:1605.04328 [physics.chem-ph] (2016).

Multi-Scale Modelling of Nanoparticle Suspensions Pietro Asinari[1], Annalisa Cardellini, Matteo Alberghini, Matteo Fasano, Eliodoro Chiavazzo Politecnico di Torino, Department of Energy, Multi-Scale Modelling Lab (SMaLL) Politecnico di Torino, Department of Energy, Multi-Scale Modelling Lab (SMaLL)

Self-assembly of nanoparticles (NPs) into mesoscopic ordered structures plays a crucial role in a large variety of applications including pharmaceutical, food, drug delivery, immunology and technological. On the one hand, trying to prevent and avoid the self-organization of nanoparticles has traditionally been the main issue for stabilizing nano-suspensions, foams and emulsions. On the other hand, the aggregation of building-blocks into mesoscopic structures has allowed to explore new materials with desired functionalities and properties. For example, many experiments and some theoretical studies have shown that the chain-forming morphologies in nano-suspensions allow an enhancement of thermal properties [1]. However, due to the challenges of controlling the multiscale phenomena occurring in nano-suspensions, clear guidelines for their rational design are still missing. Despite a wide range of experimental observations, there is an increasing need to establish rigorous modelling techniques, able to explore and describe the multiscale nature of nano-suspensions [2].In the present work a multiscale modelling approach is implemented to relate the nanoscale phenomena to the macroscopic bulk properties of nano-suspensions. Specifically, Molecular Dynamics (MD) simulations and Brownian Dynamics (BD) are synergistically integrated to understand the mechanisms driving the building-block interactions and hence to predict the shapes of assembled clusters. First, the pair Potential of Mean Forces (pPMF) is computed between atomistic modelled NPs dispersed in aqueous solutions. A sensitivity analysis is carried out by altering the hydrophilicity of the nanoparticles, their surface charge and the salt concentration of the bulk solutions. The role of anionic (Sodium Dodecyl Sulfate -SDS-) and cationic (Dodecyl Trimethyl Ammomium -DTAB-) surfactants is also investigated. Second, Brownian Dynamics simulations are carried out to understand how nanoscale phenomena, like the hydration layer or steric interactions, affect the mesoscale dynamics. The Coarse Grained procedure here suggested offers a practical multiscale approach for guiding a robust and optimal design of nanoparticle suspensions.

1. Cardellini, A., Fasano, M., Bigdeli, M. B., Chiavazzo, E., & Asinari, P. (2016). Thermal transport phenomena in nanoparticle suspensions. Journal of Physics: Condensed Matter, 28(48), 483003. 2. Batista, C. A. S., Larson, R. G., & Kotov, N. A. (2015). Nonadditivity of nanoparticle interactions. Science, 350(6257), 1242477.

From the nano- to the macroscale: bridging scales for the moving contact line problem Serafim Kalliadasis[1] Imperial College London

The moving contact line problem occurs when modelling one fluid replacing another as it moves along a solid surface, a situation widespread throughout industry and nature. Classically, the no- slip boundary condition at the solid substrate, a zero-thickness interface between the fluids, and motion at the three-phase contact line are incompatible - leading to the well-known shear-stress singularity. At the heart of the problem is its multiscale nature: a nanoscale region close to the solid boundary where the continuum hypothesis breaks down, must be resolved before effective macroscale parameters such as contact line friction and slip, often adopted to alleviate the singularity [1], can be obtained. In this talk we will review very recent progress made by our group, considering the problem and related physics from the nano- to macroscopic length scales. In particular, to capture nanoscale properties very close to the contact line and to establish a link to the macroscale behaviour, we employ elements from the statistical mechanics of classical fluids, namely density-functional theory (DFT) [2-5], in combination with extended Navier-Stokes-like equations. Using simple models for viscosity and no slip at the wall, we compare our computations with the Molecular Kinetic Theory model for the dynamic vs. equilibrium contact angle behaviour, by extracting the contact line friction, depending on the imposed temperature of the fluid [6]. A key fluid property captured by DFT is the fluid layering at the wall-fluid interface, which has a large effect on the shearing properties of a fluid. [Joint work with Benjamin D. Goddard (Edinburgh), Andreas Nold (MPI for Brain Research), Nikos Savva (Cardiff), David N. Sibley (Loughborough) and Peter Yatsyshin (Imperial)]

[1] D.N. Sibley, A. Nold & S. Kalliadasis, J. Fluid Mech. 764, 445-462 (2015) [2] B.D. Goddard, G.A. Pavliotis & S. Kalliadasis, SIAM Multiscale Model. Simul. 10, 633ñ663 (2012) [3] B.D. Goddard, A. Nold, N. Savva, P. Yatsyshin & S. Kalliadasis, J. Phys.: Condens. Matter 25, 035101 (2013) [4] A. Nold, D.N. Sibley, B.D. Goddard & S. Kalliadasis, Math. Model. Nat. Phenom. 10, 111-125 (2015) [5] A. Nold, D.N. Sibley, B.D. Goddard & S. Kalliadasis, Phys. Fluids 26, 072001 (2014) [6] A. Nold, PhD Thesis, Imperial College London (2016) Performance portability in coarse-grained mesoscale complex fluids Kevin Stratford[1], Alan Gray Edinburgh Parallel Computing Centre, The University of Edinburgh NVIDIA

This short presentation will review some recent efforts to address the issue of performance portability in our complex fluid code Ludwig. The code offers lattice-based hydrodynamics coupled with coarse-grained order parameters for different systems including mixtures, liquid crystals, and charge-carrying fluids. A lightweight thread abstraction is used to allow efficient execution of a single source code on both CPU-based and GPU-based machines. As lattice- based stencil operations are largely memory-bandwidth limited, bandwidth is used as a figure of merit. I will concentrate on open issues for performance portability in this framework.

Long-timescale Simulations with Accelerated Molecular Dynamics and the EXAALT package Danny Perez[1] Los Alamos National Laboratory

One of the common motivations coarse graining atomistic simulations is the inability of standard molecular dynamics (MD) to reach sufficiently long timescales. An alternative approach that can be particularly useful for systems that evolve through rare, thermally activated, events can be thought of as a coarse graining in time. Accelerated MD is a family of methods that implement different variants of this idea. I will review the current state of Accelerated MD, with an emphasis on the recently introduced Parallel Trajectory Splicing (ParSplice). I will then briefly describe the implementation of these methods in the EXAALT software package that combines ParSplice, the LAMMPS MD code, and the LATTE density functional tight binding code. This combination of software tools deployed on large-scale computers will enable simulations over a very large portion of the time/size/accuracy simulation space and will also allow for acceleration of multiscale systems.

Systematic coarse-graining with VOTCA Denis Andrienko[1] Max Planck Institute for Polymer Research

Coarse-graining is a systematic way of reducing the number of degrees of freedom representing a system of interest. In this lecture several coarse-graining techniques, such as iterative Boltzmann inversion, force-matching, and inverse Monte Carlo will be reviewed. and illustrated by coarse-graining the SPC/E water model, liquid methanol, liquid propane, and a single molecule of hexane.

J. Chem. Theory Comput., 2009, 5 (12), pp 3211-3223

Top-down modeling of the structure formation of copolymer materials Marcus Müller[1] Georg-August University, Göttingen

Soft, coarse-grained models that only include the relevant interactions via simple potentials are useful to study the structure formation of copolymer materials on large time and length scales. The strength of the interactions can be parameterized by a few, experimentally accessible parameter like the molecular size and the incompatibility between the blocks and the softness of the interactions allows for a high molecular density, a large degree of coarse-graining and a representation of experimentally large, invariant degrees of polymerization. The talk will exemplify the advantages and challenges of such a highly coarse-grained description by discussing static properties of spatially inhomogeneous systems as well as the dynamics, and illustrate how to relate these soft, coarse-grained models to field-theoretic approaches. Flow in heterogeneous capillary networks: from imbibition to drug delivery Heiko Rieger[1] Universitat des Saarlanes

We report two modeling studies of liquid flow in heterogeneous capillary networks: the first focuses on imbibition in nanoporous Vycor glass, which is a silica substrate with an interconnected network of nanometer-sized, elongated pores. During spontaneous imbibition a wetting liquid is drawn into the pore network by capillary forces. With the help of a theoretical model we demonstrate that long lasting meniscus arrests at pore junctions dominate the dynamics of the imbibition front and are responsible for anomalous front broadening observed in experiments. The second study addresses blood and interstitial fluid flow in the microvasculature of growing tumors, which is very heterogeneous and drastically different from the hierarchical organization of the arterio-venous blood vessel network of healthy tissue. We use a multiscale simulation framework to show that the morphological characteristics of the tumor blood vessel network together with an increased vessel wall permeability lead to outward directed interstitial fluid flow, which impedes the delivery of large convectively transported drug molecules and nanoparticles to the target tumor cells.

DL_MESO: recent developments in mesoscale modelling Michael Seaton[1] CSE Department, STFC Daresbury Laboratory

DL_MESO is a general purpose mesoscale simulation package developed for CCP5, the UK collaborative computing project for modelling condensed phase systems, and is widely used by both academia and industry. It consists of highly scalable and parallelisable codes for two mesoscale modelling methodologies: the Lattice Boltzmann Equation (LBE) method and Dissipative Particle Dynamics (DPD). Focussing on the DPD code, recent developments have involved both the addition of new functionalities to incorporate a wider range of physical models and code optimisations for efficient running on various computing architectures. This talk will give a summary of some of these developments, including the addition of more efficient charge interaction models and code porting to heterogeneous computing platforms (e.g. Intel Xeon Phi).

Multiscale approaches in cancer research Pietro Lio[1] Cambridge University

TBC

Hybrid-Resolution Simulation of Alpha-Synuclein Fibrils on Membrane Surface: Compromisation or Complementation? Liang Xu[1], Damien Thompson University of Limerick University of Limerick

Hybrid-Resolution Simulation of Alpha-Synuclein Fibrils on Membrane Surface: Compromisation or Complementation? Liang Xu [1] AND Damien Thompson [1] [1] University of Limerick, Limerick, Ireland. Abstract We used hybrid-resolution molecular dynamics simulations to investigate the interactions of alpha-synuclein fibrils with model lipid bilayer membrane. The hybrid-resolution PACE force field was applied, in which the fibril structure was represented by united-atom model, whereas the lipid bilayer and water were represented by MARTINI coarse grained models. The fully atomic system will contain about 540,000 atoms, but decreases to about 140,000 particles. Also the integration time step increases from 2 fs to 4-5 fs. The PACE parameters used here enable us to examine the secondary structure change along conformational change as the secondary structure was fixed in classic MARTINI coarse grained models. Our results show that different fibril strains exhibit different binding affinity to negatively charged membrane. It needs further study to show if such difference reveals alpha-synuclein strain-dependent interactions with membrane or arises from the limitations of the hybrid- resolution force fields that bias the present simulation results. Coarse-Grained Modeling of Protein-Nanoparticle Interactions Stefano Poggio[1], Hender Lopez, Vladimir Lobaskin School of Physics, University College Dublin, Belfield, Dublin 4, Ireland University College Dublin

The increased use of nanoparticles (NP) and nanomaterials is pushing scientific research into trying to understand the mechanisms governing interactions between biomolecules and inorganic materials. It is known that, once an NP, is in contact with a biological medium, a protein corona forms on its surface [1], and that the nature of the corona is what regulates the interaction between the NP and the other biomolecules. We aim to construct a coarse-grained model for the interaction of an arbitrary protein with several inorganic nanoparticles such as gold (Au), TiO2, CdS and carbon nanotubes. The protein in this model is represented by one bead per aminoacid, and interaction of each aminoacid with the nanoparticle surface is described by a potential of mean force from atomistic simulations [2]. The 3D structure of the protein is obtained from the corresponding code in the Protein Data Bank (PDB). The nanoparticle is represented by a two-layer model, where the surface beads interact with aminoacid-bead potential that is being obtained by fitting from all-atomistic calculations, while the core interacts with the aminoacids via van der Waals forces calculated using the Lifshitz theory and is represented by a single bead. This model can describe the protein adsorption energies on the surface of NPs of arbitrary size and of any material available, provided an all-atomistic PMF of aminoacids with flat surface of the material is available and the refractive indices for modeling the protein-NP core attraction are known. We calculate protein adsorption energies for representative proteins such as α-1-antitrypsin, lysozyme, human serum albumin and nanomaterials (Au and TiO2) for NP of different size. We discuss the implications of the variation of the attraction energy with NP size for the corona content.Funding: H2020 grant SmartNanotox, contract No. 686098

[1] M. Rahman, S. Laurent, N. Tawil, L. Yahia, and M. Mahmoudi. Protein-Nanoparticle Interactions, volume 15 of Springer Series in Biophysics. Springer-Verlag, 2013 [2] E. Brandt and A. P. Lyubartsev. Molecular dynamics simulations of adsorption of amino acid side chain analogues and a titanium binding peptide on the TiO2 (100) surface. J. Phys. Chem. C, 119:18126, 2015. 4 Posters

Testing an explicitly polarizable mesoscale (DPD) model for water Silvia Chiacchiera[1], Michael Seaton [1], Andrew Masters [2], Patrick Warren [3] [1] STFC, Daresbury, United Kingdom [2] Manchester University, United Kingdom [3] Unilever R&D, Port Sunlight, United Kingdom

We consider here a simple model for an explicitly polarizable solvent, to be used in Dissipative Particle Dynamics (DPD) simulations, and test it in various settings. The solvent is made of harmonic dimers, formed by oppositely charged beads (+q, -q), and we have access to its electric permittivity (from zero field charge dipole fluctuations, via linear response theory, LRT). We analyse various physical situations to confirm that the solventbehaves as expected, i.e., as a medium having the electrical permittivity predicted by LRT. Our tests of the dimer solvent include: the direct response to an external electric field,the reduction of the force between two ions embedded in a dielectric, the polarization charge around a ion, and the ion distribution close to a oil-water interface, to assess image charge effects. The results are obtained with the DPD code from the DL_MESO packagefor mesoscale simulations of fluids.

Charge effect on the stability of micelles in polymer/surfactant solutions: dissipative particle dynamics simulation Anastasia Markina[1], Alexander Buglakov[1], Alexey Gavrilov[1], Viktor Ivanov[1], Pavel Komarov[2][3] [1] Moscow State University, Moscow, Russia [2]Tver State University, Tver, Russia [3]Institute of Organoelement Compounds RAS, Moscow, Russia

Dissolution and precipitation of polymers in different solvents is an important field of interest in polymer science because it plays the key role in many technological process in the industry, such as fiber production, surface covering, environmental protection, drug delivery, etc. This presentation provides a model to describe a polymer film formation as a result of precipitation of polymeric micelles stabilized by a surfactant.We have developed a coarse-grained model for the aqueous solution of an acrylic copolymer and a surfactant. We use the method of dissipative particle dynamics [1-3]. Electric charges on surfactant molecules and counter-ions are taken into account explicitly, and we use for electrostatic interactions the computational scheme developed in Ref.[2]. The aim is to study the structure and to perform the analysis of stability of micelles in polymer/surfactant solutions, as well as to search for optimal conditions of a film formation from aqueous solutions of polymer micelles stabilized by surfactants. To model the film formation we apply an algorithm for evaporation of the solvent (water). As an illustration of the capabilities of the computational scheme, we demonstrate our results of the influence of various parameters of the model (concentration of components, parameters of volume interaction and of electrostatic interaction) on the characteristics of the emerging films.Mixed micelles of polymer with uncharged surfactant tend to aggregate (as well as uncharged colloids). Systems with charged surfactant molecules can behave differently depending on the strength of the electrostatic interaction. Both with a weak electrostatic contribution and with a strong contribution, the small micelles are unstable, coagulate into micelles of larger size, and form uniform films upon solvent evaporation. If we want to stabilize the small size of micelles in the emulsion, we need to find the optimal value of charges (for example, the optimum value of added salt).

[1] R. Groot and P. Warren, J. Chem. Phys. 107 4423 (1997). [2] A. Gavrilov, A. Chertovich and E. Kramarenko, Macromolecules 49 1103 (2016). [3] P. Español and P. Warren, J. Chem. Phys. 146 150901 (2017). Coarse-Grained Modeling of Protein-Nanoparticle Interactions Stefano Poggio[1], Hender Lopez, Vladimir Lobaskin University College Dublin

[1] M. Rahman, S. Laurent, N. Tawil, L. Yahia, and M. Mahmoudi. Protein-Nanoparticle Interactions, volume 15 of Springer Series in Biophysics. Springer-Verlag, 2013 [2] H. Lopez, V. Lobaskin, J. Chem. Phys.,143:243138, 2015.

Coarse Graining of Human Serum Albumin David Power[1], Stefano Poggio, Hender Lopez, Vladimir Lobaskin University College Dublin University College Dublin

In this poster we give an overview of the methods used in a new protein coarse graining technique. The procedure is a two step process, the reduction of constituent protein components and the generation of unique potentials which preserve the proteins absorption characteristics. The reduction of components was done using the CG Tool suite packaged with VMD. The potentials were generated using a genetic algorithm minimization approach with reference to a more complicated model. We were able to reduce the protein human serum albumin to a representation with only 11 beads.

Designing Uni-Molecular FRET Probes Using Multiscale Simulations Shourjya Sanyal[1], Donal Mackernan University College Dublin, Dublin, Ireland University College Dublin, Dublin, Ireland

Biosensors based on unimolecular resonance energy transfer are widely used to elucidate spatiotemporal processes within, and between cells to study transport and signalling. Improving their sensitivity is of considerable interest. The sensors, which themselves are protein complexes, are designed to measure changes in the local concentration of target molecules, ligands or ions which can be visualised and tracked using optical microscopy in vivo, and in vitro. Often, such sensors are composed of a ligand binding domain connected by a linker to a sensor domain, with a fluorescent protein covalently incorporated on either end. The choice of linker can be important and has motivated recent experimental work aimed at optimisation of flexible linker design.The studies presented here use simulation to reveal features that can be difficult tointerpret from experiments. For example, with the aid of our simulation models wepresent a robust method for extracting details of the attractive interactions betweenthe ligand binding and sensor domains using FRET microscopy data. Our simulation studies also enable us to identify the factors that can influence the signal-to-noise ratio, and from this knowledge we can begin to propose improved biosensor designs. To this end our studies include models where the flexible linkers are replaced by simple hinges.

Many-body effects in systematic coarse graining Christoph Scherer[1], Dennis Adrienko Max Planck Institute for Polymer Research Max Planck Institute for Polymer Research

Dr Liang Xu[1], Damien Thompson University of Limerick University of Limerick

Hybrid-resolution simulation of alpha-synuclein fibrils on membrane surface: Compromisation or complementation? Liang Xu [1] AND Damien Thompson [1] [1] University of Limerick, Limerick, Ireland. Abstract We used hybrid-resolution molecular dynamics simulations to investigate the interactions of alpha-synuclein fibrils with model lipid bilayer membrane. The hybrid-resolution PACE force field was applied, in which the fibril structure was represented by united-atom model, whereas the lipid bilayer and water were represented by MARTINI coarse grained models. The fully atomic system will contain about 540,000 atoms, but decreases to about 140,000 particles. Also the integration time step increases from 2 fs to 4-5 fs. The PACE parameters used here enable us to examine the secondary structure change along conformational change as the secondary structure was fixed in classic MARTINI coarse grained models. Our results show that different fibril strains exhibit different binding affinity to negatively charged membrane. It needs further study to show if such difference reveals alpha-synuclein strain-dependent interactions with membrane or arises from the limitations of the hybrid- resolution force fields that bias the present simulation results. 5 Participant List

Organizers Lobaskin, Vladimir University College Dublin, Ireland Mac Kernan, Donal University College Dublin, Ireland Pagonabarraga, Ignacio CECAM EPFL, Switzerland

Andrienko, Denis - Max Planck Institute for Polymer Research, Germany Asinari, Pietro - Politecnico di Torino, Italy Buchete, Nicolae-Viorel - University College Dublin, Ireland Casciola, Carlo Massimo - University of Rome I ''La Sapienza'', Italy Cazade, Pierre - University if Limerick, Ireland Cheung, David - NUI Galway, Ireland Chiacchiera, Silvia - STFC, United Kingdom Donev, Aleksandar - Courant Institute, New York University, USA Duenweg, Burkhard - Max Planck Institute for Polymer Research, Mainz, Germany English, Niall - University College Dublin, Ireland Español, Pep - National University of Distance Education, Spain Harb, Moussab - KAUST University, Saudi Arabia Holm, Christian - Institute for Computational Physics, University of Stuttgart, Germany Jung, Gerhard - Institute of Physics - Johannes Gutenberg Universität Mainz, Germany Kabedev, Aleksei - University College Dublin, Ireland Kalliadasis, Serafim - Imperial College, United Kingdom Kobayashi, Hideki - Max Planck Institute for Polymer Research, Germany Kremer, Kurt - Max Planck Institut for Polymer Research, Mainz, Germany Lemmin, Thomas - UCSF, USA Lio, Pietro - University of Cambridge, United Kingdom Lopez, Hender - University College Dublin, Ireland Lyubartsev, Alexander - Stockholm University, Sweden Markina, Anastasia - Faculty of Physics, Lomonosov Moscow State University, Russian Federation McCartan, Sarah - University College Dublin, Ireland Müller, Marcus - Georg-August University, Göttingen, Germany O'Reilly, Eoin - Tyndall National Institute at University College Cork, Ireland Perez, Danny - Los Alamos National Laboratory, USA Poggio, Stefano - University College Dublin, Ireland Power, David - University College Dublin, Ireland Praprotnik, Matej - National Institute of Chemistry, Ljubljana, Slovenia Reese, Jason - University of Edinburgh, United Kingdom Rieger, Heiko - Saarland University, Germany Samantray, Suman - National University of Ireland, Galway, Ireland Sanyal, Shourjya - CASL, School of Physics, UCD , Ireland Scherer, Christoph - Max Planck Institute for Polymer Research, Germany Seaton, Michael - Science & Technology Facilities Council, United Kingdom Sevink, Agur - Leiden University, The Netherlands Stratford, Kevin - University of Edinburgh, United Kingdom Tywoniuk, Bartlomiej - University College Dublin, Ireland Xu, Liang - University of Limerick, Ireland Zavadlav, Julija - ETH Zurich, Switzerland