CONICET | Buscador de Institutos y Recursos Humanos

Crystalstructure prediction (CSP) is an area of increasing scientific importance [1,2]. The development of effective computational methods for CSP wouldpotentially lead to an increase in the understanding of crystal growth andstructural analyses. In a more applied sense, CSP has importance in thepharmaceutical industry for detecting crystal polymorphs and formulating co-crystalsto improve bioavailability of drugs. Over the past ten years great strides havebeen made showing that crystal structure prediction is viable and practical.Especially promising are recent advances using dispersion corrected densityfunctional theory (DFT-D) in obtaining accurate predictions [3].At the heart of CSP methods there are twofundamental tasks. The first one is the energetic ranking of structures In mostcases a molecule will crystallize in an energetically stable form, consequently,the accurate assessment of lattice energies is essential in crystal structureprediction. Of particular importance is the balance between intra- andinter-molecular forces; correctly representing this interplay is critical. Thesecond task is efficiently exploring the global energy hypersurface, as theenergy landscape of an arbitrary molecular crystal can be quite complex,containing many local energy potential wells. Properly searching this surfacerequires a thorough but efficient search algorithm to find the global minimum aswell as other local minima in close energetic proximity, which may exist in thecase of polymorphism. The ModifiedGenetic Algorithm for Clusters and Crystals (MGAC) [4?6] is a parallel distributed, multithreaded application for crystalstructure prediction of small organic molecules. The algorithm uses populationbased natural selection to perform the global search of the energyhypersurface. Most recently we have shown that DFT-D is effective when appliedas the sole energy ranking method for MGAC [7]. By integrating the Quantum Espresso software suite [8] with MGAC we have been able to correctly predict the three atmosphericpressure polymorphs of glycine [3] and the single crystal form of histamine (unpublished results),when performing searches in the native space groups. The primary drawback ofusing DFT-D for CSP is the substantial computational time required. When compared with other commonly used methodsfor energetic ranking in CSP (such as molecular mechanics), DFT-D iscomputationally more expensive by a factor of at least 1000 when performinglocal optimization. In light of thisfact, to perform blind predictions of crystal structures of molecules ofpharmaceutical interest a significant scale up of the resources that MGAC canefficiently utilize will be required. The current version of MGAC relies on a server/clientmethod which does not scale well for large node counts (greater than 100), nordoes it allow for on the fly workload redistribution. Relevant to this is theinefficiency of the scheduling: when scaling to large numbers of nodes MGACwill often leave nodes idle for extended periods of time, leading to efficiencyas low as 50% in some cases. MGAC is also highly susceptible to file systeminstabilities, and does not have a robust mechanism for dealing with nodefailures or sudden job termination. To mitigate these factors we have begun thedevelopment of MGAC2, which is projected to be functional and in production bymid-April, 2015. By mid-May we expect to have validated MGAC2 against glycineand histamine, results that we expect will be presented at this conference.As part of thedevelopment of MGAC2 we will be relying heavily on XSEDE to provide theresources required to perform blind CSP searches. We intend to show that blindCSP searches using DFT-D are both tenable and effective in finding crystalstructures of pharmaceutical interest. We also intend to show that large scalesimulations of population based systems can be scaled effectively to largenumbers of cores, in a robust way that maximizes resources. Finally, we willpresent benchmarks for various molecules of pharmaceutical interest, as wellpreliminary results for the upcoming sixth blind test hosted by the CambridgeCrystallographic Data Centre.[1] G.M. Day, T.G. Cooper, A.J. Cruz-Cabeza, K.E. Hejczyk, H.L.Ammon, S.X.M. Boerrigter, et al., Significant progress in predicting thecrystal structures of small organic molecules - a report on the fourth blindtest, Acta Crystallogr. Sect. B-Structural Sci. 65 (2009) 107?125.[2] BardwellDA, Adjiman CS, Ammon HL, Arnautova YA, Bartashevich E, Boerrigter SXM, et al.,Towards crystal structure prediction of complex organic molecules - a report onthe fifth blind test, Acta Cryst. B67 (2011).[3] A.M.Lund, G.I. Pagola, A.M. Orendt, M.B. Ferraro, J.C. Facelli, Crystal structureprediction from first principles: The crystal structures of glycine, Chem.Phys. Lett. 626 (2015) 20?24.[4] S.Kim, A.M. Orendt, M.B. Ferraro, J.C. Facelli, Crystal Structure Prediction ofFlexible Molecules Using Parallel Geneic Algorithms with Standard Force Field,J. Comp. Chem. 30 (2009) 1973?1985.[5] V.E.Bazterra, M.B. Ferraro, J.C. Facelli, Modified genetic algorithm to modelcrystal structures. I. Benzene, naphthalene and anthracene, J. Chem. Phys. 116(2002) 5984?5991.[6] V.E.Bazterra, M.B. Ferraro, J.C. Facelli, Modified genetic algorithm to modelcrystal structures. II. Determination of a polymorphic structure of benzeneusing enthalpy minimization, J. Chem. Phys. 116 (2002) 5992?5995. [7] A.M.Lund, A.M. Orendt, G.I. Pagola, M.B. Ferraro, J.C. Facelli, Optimization ofCrystal Structures of Archetypical Pharmaceutical Compounds: A Plane-Wave DFT-DStudy Using Quantum Espresso, Cryst. Growth Des. 13 (2013) 2181?2189.[8] P.Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, et al.,QUANTUM ESPRESSO: a modular and open-source software project for quantumsimulations of materials., J. Phys. Condens. Matter. 21 (2009) 395502.