NANO-D: Modeling and Simulation of Nano-Bio Systems

Please download to get full document.

View again

of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 4 | Pages: 21

Extension: PDF | Download: 0

Related documents
NANO-D: Modeling and Simulation of Nano-Bio Systems Team Proposal (Summary) Stéphane Redon INRIA Rhône-Alpes 655 avenue de l Europe Saint-Ismier Cedex France 1 Research program
NANO-D: Modeling and Simulation of Nano-Bio Systems Team Proposal (Summary) Stéphane Redon INRIA Rhône-Alpes 655 avenue de l Europe Saint-Ismier Cedex France 1 Research program (summary) 1.1 Introduction Among the numerous potential applications of nanotechnology, one of the most promising is biology in general, and health care in particular. Indeed, besides bio-inspired nano-materials that self assemble, the increasing possibility of controlling matter at the atomic scale allows us to consider designing complex nano-systems super-drugs mono-molecular automata which would interact in sophisticated ways with biological macromolecules for therapeutic purposes. For many reasons, it is necessary to have a rationalized procedure to design these nano-systems. This may be simply to minimize costs of experimentation and production of these nano-systems, but also to quantitatively estimate a series of characteristics related to their function and relationship to their environment in particular, their potential toxicity. For example, a nano-drug would be evaluated according to criteria used in rational drug design (ADME Absorption, Distribution, Metabolism, and Excretion). In order to rationally design these nano-systems, however, we must be able to model and simulate them and their environment. For several reasons, this is extremely costly: the number of involved atoms is extremely large (liposomes, proteins, viruses, DNA, cell membrane, etc.); some biological phenomena have large durations (e.g. the folding of some proteins); finally, the underlying physico-chemistry of some phenomena can only be described by quantum chemistry (local chemical reactions, isomerizations, metallic atoms, etc.). For all these reasons, modeling and simulating nano-bio phenomena usually requires powerful and costly computing resources such as super-computers, computer clusters, grids, etc. This constitutes a major impediment to the development of molecular nanobiotechnology. Of course, one possible strategy would be to wait for computer processing power to increase, but this strategy is not always acceptable (for example in health). If nanotechnology is to have a significant impact, it will thus be necessary to develop powerful enough computational methods. Our goal is to go beyond the current state of the art in modeling and simulation of nano-bio systems by developing a set of adaptive methods for macro-molecular and quantum mechanics. These methods will be able to automatically and rigorously focus computational resources on the most relevant parts of a simulated molecular system, and will allow for tight expert-in-the-loop analysis and design of nano-bio systems. Our theoretical and computational methods will be experimentally validated through collaborations with chemists, bio-physicists and biologists. We envision several software tools based on our methods which will allow for rapid design and prototyping of complex nano-bio systems (similar, for example, to what CATIA by Dassault-Systèmes allows for macroscopic 1 2 Proposed team 2 objects). These tools will allow a user to model and simulate both nano and bio objects, as well as their interactions. We believe these tools will have numerous applications, including design of nano-materials, biosensors, nano-drug design, study of quantum phenomena in biology, wet nano-electronics, etc. Furthermore, because the methods we will design will adapt to available computing resources, they will be usable on a wide variety of systems, in particular low-end workstations with widespread availability. 2 Proposed team The composition of the proposed team is the following: Stephane Redon. CR1 INRIA. Sergei Grudinin. Post-doctorant INRIA. Serge Crouzy. CEA DSV, HDR, collaborateur extérieur. Karthik Arumugan. Doctorant CEA (co-encadrement CEA). The resumes of Stephane Redon, Sergei Grudinin and Serge Crouzy are attached at the end of this document. 3 Strategic considerations Impact: the wide applicability of molecular modeling and simulation gives us the potential to have a significant impact. Furthermore, all algorithms and tools we know of require clusters of processors, sometimes super-computers, in order to manipulate complex molecular systems while retaining some amount of detail. By focusing first on the development of adaptive algorithms, we see two major benefits. First, we have a chance to be among the first to develop such algorithms and open new research directions. Second, by developing algorithms that are also suitable for low-end computers, by nature largely available, we have a chance to have a significant impact in the targeted research communities (nano-materials, nano-objects, nano-design, nano-biotechnologies, and biology). Collaborations: researching basic algorithms should help us attract a large number of collaborators, both academic and industrial. Our goal is to develop and maintain multidisciplinary collaborations involving computer scientists as well as physicists and biologists (as we have begun, see below), in order to help us (a) guide our research and ensure its relevance, (b) validate our research and (c) disseminate our research and the tools we will develop. INRIA Rhône-Alpes is ideally suited for such a purpose (see below). Team building: we believe the multidisciplinary abilities required by the subject (computer science, mathematics, physics, chemistry and biology), the presence of difficult problems with potentially broad impact, as well as the general excitement surrounding research in nano- and biotechnologies, should greatly help us attract and recruit high-quality students and team members. Funding: nano- and bio-technologies are strategic research axes within funding programs at all levels. Researching basic algorithms for modeling and simulation allows us to apply for grants in several programs, with multidisciplinary teams. Furthermore, being able to develop software tools based on our research allows us to apply for funding in programs concerned with industrial transfer as well (see below). For example, relevant ANR programs in 2008 are Conception et Simulation, Contenus et interactions, PNANO, Blanc, and Jeune chercheur (although at present the Nano-D team is too small to participate in the Jeune chercheur program). Dissemination and transfer: we are currently in the process of patenting the methods we have developed (the adaptive force update algorithms, which are at the core of the adaptive simulation algorithms), 4 Location 3 and we have recently applied for funding to help initiate industrial transfer (in the GRAVIT Focus Innovation funding program). We believe a technology enabling virtual prototyping of molecular systems may have numerous industrial applications (and increasingly more so), and there may be some opportunities to provide licences or to transfer technologies to start-ups in the near future. Meanwhile, we will distribute the software prototypes we will develop (in executable form). This will help us (a) establish the team and communicate on our research, (b) ensure the research is relevant (through user feedback), and (c) develop collaborations. The targeted publications will consist of conference proceedings (ISMB, ECCB, RECOMB, NanoTech, etc.) and journals (Bioinformatics, Journal of Computational Chemistry, Journal of Computational Biology, Journal of Computational Physics, etc.). When appropriate, we may also submit papers in related fields (e.g. robotics and graphics: RSS, ICRA, IROS, SIGGRAPH). Through collaboration with biologists and physicists, we might also participate in submissions to natural sciences journals (e.g. PNAS, Nature, Nature Nanotechnology, etc.). 4 Location The INRIA Grenoble - Rhône-Alpes Research Center is an ideal location to develop a team focusing on modeling and simulation of nano-bio systems, thanks to the incredibly rich research environment on both nanotechnology and biology. The region enjoys the presence of many research organisms, labs, facilities, networks, universities with specialized courses, including for example: CEA: CEA hosts several research teams in nanoscience and biology. Our main partner in the one-year ANR AMUSIBIO project, Serge Crouzy, leads the MIR group at CEA. Minatec: Minatec groups 2000 researchers and professors in micro/nanotechnologies. Minalogic: Minalogic is a world competitiveness cluster ( Pôle de compétitivité mondial ) on micro/nanotechnologies and embedded software, which groups public and private research centers (Minatec, INPG). European Molecular Biology Laboratory (EMBL): one facility of the EMBL is located in Grenoble (out of five facilities, with Hambourg, Heidelberg (main laboratory), Hinxton, and Monterotondo). The Grenoble unit hosts seven research groups. European Synchrotron Radiation Facility (ESRF): a collaboration between eighteen European countries, the ESRF is located in Grenoble. Institut Laue-Langevin (ILL): ILL, in Grenoble, is the world s leading facility in neutron science and technology. Institute of Structural Biology (ISB): located in Grenoble, the ISB hosts twelve groups. Lyonbiopôle: Lyonbiopôle is a world competitiveness cluster ( Pôle de compétitivité mondial ) on vaccines and diagnostics. RTRA Nanosciences Grenoble: Réseau Technologique De Recherche Avancée (Technological Network of Advanced Research). One of thirteen excellence networks. This network groups CEA, CNRS, INP Grenoble and UJF, totalling 600 researchers. Eight main research themes are being investigated. Quantum nanoelectronics Nanomagnetism and spin electronics Nanophotonics Molecular electronics Nanomaterials, nanobonding, nanostructuring Nanocharacterization et nanometrology 5 Current collaborations and contacts 4 Life and nanoelectronics Nanomodeling theory and simulation Industries: several large companies, including biomérieux, Sanofi Pasteur, etc. Students: The Rhône-Alpes region hosts many universities, engineering schools, etc., including some specialized trainings in nanotechnology (for example the N3 Master at Joseph Fourier University: Nanosciences, Nanomatériaux, Nanotechnologies ). 5 Current collaborations and contacts We have already developed several contacts: Serge Crouzy, CEA, Grenoble: Serge Crouzy, who leads the MIR group at CEA, is a bio-physicist who co-developed the adaptive molecular dynamics algorithms with us during the one-year ANR AMUSIBIO project. Michel Vivaudou, ISB, Grenoble: Michel Vivaudou, formerly in CEA, is now a member of the Laboratory of Membrane Proteins in ISB, headed by Eva Pebay-Peyroula. His research deals with ABC transporters, and he was involved in the ANR AMUSIBIO project to help evaluate the tools developed during the project. Martin Field, ISB, Grenoble: Martin Field, who leads the molecular dynamics laboratory at ISB, has a well-recognized expertise on theoretical and computational development of mixed quantum mechanics/molecular mechanics. Jean-Michel Jault, ISB, Grenoble: Jean-Michel Jault is a biologist who heads the Jeunes Chercheurs team on ATPases et GTPases bactériennes. He is a specialist of ABC transporters. Christian Joachim, CEMES, Toulouse: Christian Joachim leads the GNS, a well-known research group in nano-science which has pioneered a number of theoretical and practical approaches enabling the development of nano-systems (Christian Joachim has been awarded several prizes for his research, including the 2001 CNRS Silver Medal in Chemistry for his work on molecular nanosciences, and the 2005 Feynman Prize in Theory for his design of mono-molecular devices). Michael Nilges, Institut Pasteur, Paris: Michael Nilges leads the Structural Bioinformatics unit at Institut Pasteur in Paris. One of the focus of the group is the development of computational methods to determine protein structure from NMR data. Konrad Hinsen, CBM, Orléans: Konrad Hinsen is a researcher in the Center of Molecular Biophysics ( Centre de Biophysique Moléculaire ) in Orléans and associated researcher at the Synchrotron Soleil in Saint Aubin since The current research activities of the group (led by Gerald Kneller) are focused on the study of slow protein dynamics by computer simulation, in particular the development of stochastic simulation techniques, models for protein dynamics, and parallel computing. Bernard R. Brooks, NIH, Bethesda: Bernard R. Brooks leads the Laboratory of Computational Biology at NIH, Bethesda, and is one of the main contributors of CHARMM, a well-known molecular dynamics software package with a very large user base. We recently met Bernard Brooks at the first NIH-INRIA Workshop in Bethesda, and he offered to fund visits to his lab in order to collaborate on molecular dynamics simulation, as well as interface the tools we developed during the ANR AMUSIBIO project with the CHARMM package. 6 Positioning within INRIA 5 6 Positioning within INRIA Modeling and simulation of nano-bio systems is complementary to the activities of a few INRIA teams, and we would like to develop collaborations with them as well: ABS: ABS develops methods for computational structural biology. Inspired by computational geometry, the group focuses on modeling interfaces and contacts between bio-molecules, the flexibility of macro-molecules, as well as shape learning and applications. These geometric aspects, centered on biomolecules, are complementary to our dynamics and robotics approach. We also have a partial overlap of applications, since structural biology is a potential application of computational tools for modeling and simulation of nano-bio systems. MICMAC: MICMAC, is a joint project-team of INRIA and ENPC. The group focuses on theoretical and numerical issues in quantum chemistry (electronic structure calculations), molecular dynamics problems (computation of free energy differences, sampling methods, etc.) and multiscale problems (including multiscale modeling of solids and fluids). We believe a collaboration with MICMAC on theoretical problems, especially on adaptive molecular dynamics and sampling methods, could be particularly helpful. ScAlApplix: ScAlApplix is a group focusing on high performance parallel computing. One application studied in the group is molecular chemistry, a problem which is in theory extremely scalable. If possible, we believe a collaboration with several other teams from the INRIA Grenoble - Rhône-Alpes Research Center could be fruitful, including for example Artis (on methods for rendering complex molecular datasets), BIPOP (on theoretical and numerical methods for energy minimization), emotion (robotic path planning methods have been used in the past to determine molecular pathways - this is also one research theme of the robotics group headed by Jean-Paul Laumond in Toulouse), Evasion (on the development of multiscale methods, and user interfaces), i3d (on 3D interaction and haptic rendering with complex molecular data), MOAIS (on distributed computing applied to interactive molecular design), and MESCAL (on parallel computing). 7 Summary of goals Our main short-term (about two years) scientific objectives are: Modeling complex molecular models. One fundamental problem when modeling complex molecular systems is essentially to provide the user with intuitive control of the numerous degrees of freedom in the system. Unfortunately, intuitive interaction is difficult to achieve because it is non-trivial to map a user interface, which has only a few degrees of freedom, to the molecular system which may have several thousands of degrees of freedom. Based for example on our hierarchical representation of a molecular system, we will develop multilevel interaction methods which will allow the user to easily perform complex, large-scale structural modifications of macromolecular systems. Adaptive implicit solvation. Solvation is important in many biological phenomena. Unfortunately fast, explicit simulation of solvant (that is, representing all solvent molecules explicitly) is difficult because of the numerous degrees of freedom involved. To address this problem, implicit solvation methods have been developed. Sergei Grudinin, a post-doctoral researcher with a mixed bio-physics and computer science background, has been developing such an implicit solvant method based on a boundary element approach. We will generalize this approach so that it can be used in our adaptive dynamics framework. Fast approximate force fields. Classical (i.e. Newtonian) molecular mechanics relies on the periodic evaluation of the inter-atomic force field. Due to the quadratic complexity of this operation (when cutoff distances are not used), simulation of macro-molecular systems on long time intervals are extremely difficult. Within AMUSIBIO, we have proposed an adaptive algorithm to update the intra- and extramolecular forces (van der Waals, electrostatic, dihedral angles). This algorithm exploits the rigidification 7 Summary of goals 6 of subgroups of atoms to perform a partial update of the pairwise forces, resulting in a simulation speedup. Up to now, the algorithms we have developed have concerned the update of the exact semi-empirical force fields traditionally used in classical molecular mechanics. For dense macromolecular systems (such as a protein in its native state), however, reducing the number of degrees of freedom in the system does not necessarily lead to a significant decrease of the number of pairwise forces that have to be updated, in particular because of the long range of the electrostatic interactions. Thus, we will develop complementary algorithms, inspired from approximate force field computations (e.g., the Fast Multipole Method), that will be adapted to the algorithms we have recently introduced. Adaptive quantum mechanics. Quantum mechanics is essential to the modeling of some nano-systems, when precise information about electronic structure and dynamics is required. In collaboration with Martin Field (IBS) and Thierry Deutsch (CEA), we want to develop an adaptive quantum mechanics approach, inspired by our work on adaptive Newtonian mechanics. Briefly, we will first attempt to perform a hierarchical decomposition of the Hamiltonian of the system, similar to our current hierarchical representation. Then, we will investigate how this representation can be used to incrementally update various coefficients involved in quantum mechanics computations, in order to speed up the resolution. Finally, as in our current adaptive approach, we will attempt to develop methods to estimate a priori error bounds and restrict computations to limited subsets of the molecular system. In parallel, we will attempt to set up an introductory course on molecular dynamics, and Stephane Redon will apply for the Habilitation à diriger des recherches. Stephane Redon IEEE Member, ISCB Member Date of Birth: August, 17, 1975 Web: Phone: Address: INRIA Rhône-Alpes Research Unit 655 avenue de l'europe - Montbonnot Saint Ismier Cedex - France February 19, 2008 Appointments 03/05- : INRIA: Research Scientist 12/02-03/05 : University of North Carolina at Chapel Hill: Post-Doctoral Research Associate Research Assistant in the GAMMA team, headed by Pr. Lin and Pr. Manocha ( Education 09/99-10/02 : INRIA Rocquencourt: Ph.D. Dissertation title: Algorithms for interactive dynamics simulation of rigid bodies Advisors: Dr. Sabine Coquillart and Dr. Abderrahmane Kheddar Honored with the congratulations of the board 09/98-09/99 : Paris VI: D.E.A. (Master s degree) Specialties: Computer Graphics and Cryptography Internship at INRIA-Rocquencourt with Dr. Sabine Coquillart Project: Continuous collision detection for rigid polyhedral objects 09/95-09/98 : Ecole polytechnique: Engineer degree (X 1995) 3 rd year: Specialization in Computer Graphics and Economy 2 nd year: Common-core syllabus, then specialization in Mathematics and Economy 1 st year: Military duty at ESM Saint-Cyr (assistant to the Head of Communication Services) Recent Projects and C
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks