A.T. Phillips, J.B. Rosen, and V.H. Walke, Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima, Dimacs Series in Discrete Mathematics and Theoretical Computer Science 23 (1995), P.M. Pardalos, G.-L. Xue, and D. Shalloway (Eds), 181-198.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....heterogeneous systems have emerged as a powerful platform for executing high performance applications. Current experience shows that many high performance distributed applications are composed of a few coarse grained tasks and execute on heterogeneous platforms composed of 2 or 3 machines [28, 38, 44, 49]. These platforms are formed by smart instruments and machines which can be SIMD MPPs, MIMD MPPs, vector computers, and or workstations. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used ....

....vector computers, and or workstations. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector computer is used to execute the serial portion [44]. For such applications, the development of a performance efficient allocation of tasks to machines is dependent upon a realistic prediction of application behavior and, in particular, of application computation and communication costs under changing system load. Additional applications executing ....

[Article contains additional citation context not shown here]

A. Phillips, J. Rosen, and V. Walke, "Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima", University of Minnesota Supercomputer Institute Research Report UMSI 94/126, July 1994.

....heterogeneous systems have emerged as a powerful platform for executing high performance applications. Current experience shows that many high performance distributed applications are composed of a few coarse grained tasks and execute on heterogeneous platforms composed of 2 or 3 machines [28, 38, 44, 49]. These platforms are formed by smart instruments and machines which can be SIMD MPPs, MIMD MPPs, vector computers, and or workstations. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used ....

....vector computers, and or workstations. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector computer is used to execute the serial portion [44]. For such applications, the development of a performance efficient allocation of tasks to machines is dependent upon a realistic prediction of application behavior and, in particular, of application computation and communication costs under changing system load. Additional applications executing ....

[Article contains additional citation context not shown here]

A. Phillips, J. Rosen, and V. Walke, "Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima", University of Minnesota Supercomputer Institute Research Report UMSI 94/126, July 1994.

....that cannot be performed efficiently in this environment. Many vector codes have been reworked to suit networked resources, but this can be a difficult task. Small number of tasks. Distributed parallel applications generally consist of a 17 small number of tasks [CS95, DS95, DTMH96, PRW94, MMFM93, WK93, YB95] The size of individual tasks vary from application to application. Modular Implementation. Many distributed parallel applications are actually several parallel applications joined together to solve a larger problem. For example, the Global Climate Modeling group at UCLA ....

....and to make selections 100 between them. One of the problems we have encountered is that many approaches to modeling assume the availability of pieces of data that we do not have for our system. Thus, we have grounded this work in models used for current applications [CS95, DS95, DTMH96, MMFM93, PRW94, WK93, YB95] These models all take into account complexity of computation, information availability, and accuracy needed for predictions in a parallel distributed environment. POEMS [DDH 98] is an end to end modeling environment consisting of compiler driven task graph generation and ....

A. T. Phillips, J. B. Rosen, and V. H. Walke. Molecular structure determination by convex global underestimation of local energy minima. Technical Report Tech Report UMSI 94/126, University of Minnesota Supercomputer Institute, 1994. 244

....j are both H type and 0 otherwise. As the reader may observe this function is indeed very difficult to minimize globally. In the literature there are various reports of solving this approximately for simplified models (homo polymer problems for instance where A 1 = A 2 = AN , for small N) [21]. b) The Lattice Model: The number of all possible conformations of a polypeptide chain is too large to be sampled exhaustively, yet protein sequences do fold into unique native states in milliseconds this is known as the Levinthal paradox. Lattice models have been used to address the protein ....

....2.4.1. Ring Closure: The problem is to compute conformations of a molecule with a cyclic structure in which the constraints imposed by the bond lengths and angles are respected. Motion Planning Protein Docking See [25] Robotics) Numerical Optimization Protein folding See section 2.4.1. Also see [21]. Computational Linguistics Understanding genetic sequences. See [22] Text Compression DNA Classification See [24] Techniques DNA Sequence Identification Task: It is hard to give a very precise definition of the problem. This can only be explained by giving examples: classifying bacterial ....

Philips A. T., Rosen J. B., Walke V. H., Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima, Global Minimization of Nonconvex Energy Functions: Molecular Conformation and Protein Folding, DIMACS Workshop, March 20-21, 1995.

....few coarse grained tasks and execute on heterogeneous systems formed by a few machines [1, 15, 17] These machines can be MIMD MPPs, visualization engines, SIMD MPPs, workstations, and or vector computers. For many of these applications, both heterogeneity and parallelism in the code are exploited [4, 10, 13, 14]. For instance, in a molecular structure determination application which computes the global minimum of a potential energy function [14] 1. Supported by a scholarship from CAPES and UFRJ (Brazil) 2. Supported in part by NSF contract number ASC 9301788. TR UCSD CS96 484 2 an MPP is used ....

....engines, SIMD MPPs, workstations, and or vector computers. For many of these applications, both heterogeneity and parallelism in the code are exploited [4, 10, 13, 14] For instance, in a molecular structure determination application which computes the global minimum of a potential energy function [14], 1. Supported by a scholarship from CAPES and UFRJ (Brazil) 2. Supported in part by NSF contract number ASC 9301788. TR UCSD CS96 484 2 an MPP is used to compute distinct local minima whereas a vector computer is used to compute a convex quadratic underestimator function by solving a ....

A. Phillips, J. Rosen, and V. Walke, "Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima," University of Minnesota Supercomputer Institute Research Report UMSI 94/126, July 1994.

....another depending on the specific architecture or load characteristics at run time, we allow component models to be exchanged individually as needed. Small number of sub tasks. Commonly, applications are broken into a small number of sub tasks, usually fewer than five [CS95, DS95, DTMH96, MMFM93, PRW94, WK93, YB95] This is a manageable number to model individually. In addition, we can use an application profile to help direct our efforts to model the most important components in more detail. Implementations are different. Sub tasks are implemented in different languages using different ....

....them. One of the problems we have encountered in doing this is that many approaches to modeling assume the availability of pieces of data that we do not have for our system. Thus, we have grounded our work in the models used for current applications in this area [CS95, DS95, DTMH96, MMFM93, PRW94, WK93, YB95] These models all take into account complexity of computation, information availability, and accuracy needed for predictions in this setting 4 Component Model Selection Our model selection is based on the principle of using the simple model that provides accurate results. Model ....

A. T. Phillips, J. B. Rosen, and V. H. Walke. Molecular structure determination by convex global underestimation of local energy minima. Technical Report Tech Report UMSI 94/126, University of Minnesota Supercomputer Institute, 1994.

....heterogeneous systems have emerged as a powerful platform for executing high performance applications. Current experience shows that many high performance distributed applications are composed of a few coarse grained tasks and execute on a heterogeneous platform formed by two machines [9] 13][15]. These machines can be SIMD MPPs, MIMD MPPs, workstations, and or vector computers. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector ....

....workstations, and or vector computers. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector computer is used to perform the serial portion [15]. For such applications, the development of a performance efficient allocation of tasks to machines is dependent upon a realistic prediction of application behavior under changing system load. In particular, additional applications executing on the system can dramatically affect the availability ....

[Article contains additional citation context not shown here]

A. Phillips, J. Rosen, and V. Walke, "Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima", University of Minnesota Supercomputer Institute Research Report UMSI 94/126, July 1994. TR - UCSD - CS96-477 23

....heterogeneous systems have emerged as a powerful platform for executing highperformance applications. Current experience shows that many high performance distributed applications are composed of a few coarse grained tasks and execute on a heterogeneous platform formed by two machines [9] 13][14][16] These machines can be MIMD MPPs, workstations, SIMD MPPs, vector computers, etc. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector ....

....SIMD MPPs, vector computers, etc. For many of these applications, both heterogeneity and parallelism in the code are exploited. For example, in a molecular structure application, an MPP is used for the parallel part of the code whereas a vector computer is used to perform the serial portion [14]. For such applications, the development of a performance efficient allocation of tasks to machines is dependent upon a realistic prediction of application behavior under changing system load. In particular, additional applications executing on the system can dramatically affect the availability ....

[Article contains additional citation context not shown here]

A. Phillips, J. Rosen, and V. Walke, "Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima", University of Minnesota Supercomputer Institute Research Report UMSI 94/126, July 1994.

....also been used in molecular dynamics simulations [145] An alternative method smoothes the energy hypersurface using local minima to construct separable quadratic underestimating functions. The local minima of these fitted functions are then used to initiate searches on the original energy surface [42, 142]. 3.5. Branch and Bound Methods. Typically, branch and bound methods have been associated with, and are efficient methods for, solving combinatorial global optimization problems [47, 49] The protein folding problem possesses inherently combinatorial aspects, and some explicit combinatorial ....

A. T. Phillips, J. B. Rosen and V. H. Walke, Molecular structure determination by convex global underestimation of local energy minima, In DIMACS Series in Discrete Mathematics and Theoretical Computer Science, 23, Amer. Math. Soc., (1996), 181-198.

No context found.

A.T. Phillips, J.B. Rosen, and V.H. Walke, Molecular Structure Determination by Convex Global Underestimation of Local Energy Minima, Dimacs Series in Discrete Mathematics and Theoretical Computer Science 23 (1995), P.M. Pardalos, G.-L. Xue, and D. Shalloway (Eds), 181-198.

....Lynn F. TenEyck ComputerScienceandEngineering SanDiegoSupercomputerCenter University ofCalifornia,SanDiego University ofCalifornia,SanDiego jbrosen ucsd.edu teneyckl sdsc.edu Abstract Here,weintroducea new approach tomacromoleculardockingwhich combinesthecontinuousglobaloptimizationalgorithmCGU [6,22]withthegrid basedconformationalsearch engineDOT [24] W e alsodetailtheuseofthismethod todock acetylcholineintothefasciculin acetylcholinesterase complex.Fasciculinblo cksaccesstothelong,narrow pathway leadingtotheactivesite,which islocatedintheinterior ....

....z For informationon CGU, pleasevisitthe CGU home page at: http: cgu.cs.uw ec.edu. The latterencompassesawidervariety oftechniques,includingmoleculardynamicsand statisticalmechanics[3,11,16, 23] Mon teCarloand geneticalgorithms[4,20] underestimationor tunneling methods[6,7,22]such asCGU, and proceduresforsmoothingenergylandscapes[15,18,21,26] Insteadofutilizinga singlereflnement technique,we have herecoupleda discretealgorithmwitha continuousone. Whilethisideawasarrived atindependently,itmost certainlycanbe foundintheworkofothers.W e citeasjust oneexampletheAutoDock ....

[Article contains additional citation context not shown here]

Phillips,A. T.,Rosen,J.B. and W alke,V. H. (1995), Molecularstructure determinationby convexglobalunderestimationof localenergy minima. DIMA CS Series inDiscreteMathematicsand TheoreticalComputerScience, Vol.23,GlobalMinimizationofNonconvexEnergy Functions:MolecularConformationandProteinFolding. Pardalos,P.M.,Shalloway,D.andXue,G.,Eds.AmericanMathematicalSociety, Providence,RI.Pgs.181-198.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC