Request for Proposals for Biomolecular Simulation Time on Anton

 

Introduction

The Pittsburgh Supercomputing Center (PSC, www.psc.edu) is soliciting proposals for computer time on Anton 2, a special-purpose supercomputer for molecular dynamics (MD) simulation designed by D. E. Shaw Research (DESRES). In late 2016, a 128-node Anton 2 machine will be made available without cost by DESRES for non-commercial research use by universities and other not-for-profit institutions. The machine will be hosted by PSC.

This RFP solicits proposals for simulation time on the 128-node Anton 2 machine that will be hosted at PSC, not the Anton 1 system that has been at PSC since 2010. Thus Anton in the remainder of this RFP refers specifically to the Anton 2 machine.

To qualify for an allocation on Anton, the principal investigator (PI) must be a faculty or staff member at a U.S. academic or non-profit research institution and must have the appropriate level of authority and responsibility to direct the project supported by the allocation. A graduate student or postdoctoral researcher may not be a PI, but a qualified advisor may apply for an allocation on her or his behalf. Furthermore, each investigator can serve as a PI for only a single application for computer time on Anton in a given allocation round. PIs who previously had an allocation on Anton 1 at PSC are required to provide a 2-3 page progress report of their past work on the Anton 1 machine at PSC. Repeat applications by a given PI that are unaccompanied by a progress report will not be considered.

The remainder of this Request for Proposals (RFP) describes the intended focus of solicited projects, outlines simulation requirements, offers proposal preparation instructions, and explains the proposal review process.

 

Intended Focus of Proposed Projects

Anton enables scientists to perform MD simulations of biomolecular systems nearly two orders of magnitude faster than the previous state of the art (see the Estimating Requested Simulation Resources section). To maximize the benefit of Anton to the scientific community, proposed projects should focus exclusively on questions that will be greatly advanced by long continuous MD simulations. Further, investigators should explain why multi-microsecond timescales are necessary and important for the molecular system and scientific question under consideration. Proposals to run only a large number of shorter (sub microsecond) simulations will receive a lower ranking in the review process.

 

Simulation Requirements

Anton uses specialized hardware to perform molecular dynamics simulations orders of magnitude faster than general-purpose hardware running traditional MD software (see Reference 1 below). Importantly, Anton does not run Desmond, AMBER, NAMD, GROMACS, or any other MD simulation software package, although it uses a Desmond structure (DMS) file as an initial input and its trajectory output is compatible with Desmond’s (DTR files). Each simulation must be built, using tools available at PSC, specifically for Anton.

Anton is designed primarily to accelerate classical MD simulations of biomolecular systems with periodic boundary conditions and explicit solvent. To make best use of the available computational resources, consideration will be given only to applicants whose projects satisfy the criteria outlined below. Investigators who have questions regarding the suitability of their proposed simulations are encouraged to contact This email address is being protected from spambots. You need JavaScript enabled to view it. to discuss their planned project before submitting their proposal.

 

  1. Simulations must be standard MD runs in the constant NVE, constant NVT (isothermal), or constant NPT (isothermal, isobaric) ensembles. Constant NVT and NPT simulations must use the Multigrator framework described in [2] with either Nose-Hoover or Langevin thermostats and isotropic or semi-isotropic MTK barostats. Simulation conditions may include the specification of a uniform constant applied electric field. Position restraints, on a per atom basis, are allowed. Enhanced sampling is also available in three basic forms: (i) simulated tempering with the Nose-Hoover thermostat, (ii) application of restraints between the centers of mass of groups of atoms, and (iii) application of conformational restraints, each based on the calculation of RMSD (root mean squared deviation) with respect to atomic positions of a given reference structure. For restraints in both (ii) and (iii), equilibria and spring constants can be varied during a simulation according to a schedule. Applicants with systems that have dozens of restraints and/or restraints involving thousands of atoms should contact This email address is being protected from spambots. You need JavaScript enabled to view it. before submitting a proposal.
  2. The simulation cell must have only right angles (i.e., it must be a cubic or orthorhombic box), and must be a minimum of 45 Angstroms on each side. Applicants with systems shaped such that one dimension of the simulation cell is much larger than the others should contact This email address is being protected from spambots. You need JavaScript enabled to view it. before submitting a proposal.
  3. Proposed simulations must use recent variants of the following standard, nonpolarizable biomolecular force fields: CHARMM (e.g., CHARMM22, CHARMM27 - including CMAP corrections, and CHARMM36), AMBER (e.g., AMBER99, AMBER99SB, AMBER03), or OPLS (e.g., OPLS-AA/L). Modified versions of the CHARMM and AMBER force fields, based on published research by DESRES, are also acceptable (and available through the simulation setup tools). Water should be modeled with the SPC, TIP3P, or TIP4P models, or their variants.
  4. Chemical systems proposed for simulation must contain between 25,000 and 700,000 atoms (including solvent atoms), though systems between 50,000 and 700,000 atoms are recommended for maximum efficiency. Chemical systems proposed for simulation must consist of some combination of protein, DNA, RNA, lipids, water, and standard ions. Investigators who wish to use custom parameters or molecules that are not included in the standard distribution of the supported force fields (see 3 above) should contact This email address is being protected from spambots. You need JavaScript enabled to view it. to discuss the suitability of their simulations before submitting their proposal.

 

Estimating Requested Simulation Resources

Applicants may refer to the table below, with benchmarks for a number of systems of various sizes, to estimate the amount of machine time required for their project. The actual achievable simulation times, however, may vary even for different molecular systems of similar size. No more than a total of 460,000 MD simulation units (9.1 machine-days) will be allocated to any one principal investigator. (An “MD simulation unit” is defined as the amount of machine-time required to simulate 1 nanosecond of a 50,000 atom system using Anton production parameters.) It is anticipated that 10 to 20 allocations will be made at or near 460,000 MD simulation units, and 30 to 40 allocations will be made at or near 230,000 MD simulation units. Applicants are encouraged to target their requested resources at one of these levels. Note that simulations on Anton 2 will run on all 128 nodes.

For more details regarding Anton 2, please see the References.

Chemical system (PDB ID)Number of atomsApproximate performance (microseconds/machine-day)*
DHFR (5DFR) 23,558 61.3
aSFP (1SFP) 48,423 53.0
FtsZ (1FSZ) 98,236 26.0
T7Lig (1A01) 116,650 21.9
bILAP (1BPM) 132,362 18.7
f1atpase 327,506 7.9
Tiled FDH-H** 700,184 3.6

*All simulations used 2.5-femtosecond time steps with long-range interactions evaluated at every other time step and a Nose-Hoover thermostat applied every 100 time steps. Performance was measured on a 128-node Anton 2 machine like the one hosted by PSC. Simulation performance in microseconds/machine-day is approximately 50% higher if 4fs time steps and Hydrogen Mass Repartitioning are used (see Reference 3).

**This system represents eight copies of FDH-H (formate dehydrogenase H) in a single 190 Angstrom cubical box.

 

Structure of Proposal

Proposals should be two to six pages in length, not including references. PIs who previously received an allocation on Anton 1 at PSC must also provide a separate 2-3 page progress report demonstrating their successful use of Anton 1 to produce high-impact scientific results. In addition, applicants may submit up to two additional supporting documents (e.g., published papers) in PDF format on the submission page. The main proposal document must have the following sections:

  1. Summary of the project, including descriptive title of proposed research (400 words maximum for summary & title).
  2. Name, address, email, and telephone number of Principal Investigator(s).
  3. Background information (1 page maximum). Investigators should include sufficient background information on the research field to allow reviewers to judge the scientific merit of the proposed research.
  4. Scientific Objectives to be accomplished on Anton (2 pages maximum). Investigators should clearly explain why long-timescale MD simulations are important for the planned project (see the Intended Focus of Proposed Projects section for more details). Applicants should also clearly explain the scientific impact of their proposed project.
  5. Project Feasibility (2 page maximum). Applicants must clearly and explicitly address all four points outlined in the Simulation Requirements section of this RFP, providing all necessary details regarding their proposed simulations and the expertise of their research team members (including prior experience running MD simulations) to ensure that the proposed simulations can be successfully completed on Anton.
  6. Requested Resources (1 page maximum). Investigators must clearly state and provide justification for the number of Anton MD simulation units requested for their project subject to the limits given in Estimating Requested Simulation Resources. The justification should provide strong scientific arguments as to why the length and number of proposed simulation runs will be both sufficient and necessary to achieve the stated scientific objectives. Please refer to the Estimating Requested Simulation Resources section for benchmarks to facilitate estimation of requested resources. No more than a total of 460,000 MD simulation units (9.1 machine-days) can be requested per PI.

 

Important Dates and Proposal Review Process

Proposals should be submitted electronically in PDF format at https://grants.psc.edu/cgi-bin/proposals/anton/application.pl.

The deadline for applications is 11:59 pm, EDT, Thursday, June 23, 2016.

Proposals will initially be assessed for technical feasibility by PSC staff (see the Simulation Requirements section) and will then be reviewed by a peer committee to be convened by the National Research Council. Resource allocations by PSC will follow the recommendations of the National Research Council. Resources are expected to be made available starting in November of 2016.

Proposals will be ranked based on the scientific merit of the proposed research and the strength of the justification for the requested resources. In addition, proposals by PIs that previously received an allocation on Anton 1 at PSC will need to demonstrate that the scientific outcomes from the previous award(s) justify another award. At least 25% of the total allocated time will be reserved for PIs that have not previously had an allocation on Anton 1 at PSC. Since only a limited number of allocations will be available on Anton, a successful proposal will clearly state how access to Anton will facilitate breakthrough science.

For general questions regarding this RFP, or to discuss feasibility and technical aspects of projects, please contact This email address is being protected from spambots. You need JavaScript enabled to view it..  

 

References

[1] Anton 2: Raising the Bar for Performance and Programmability in a Special- Purpose Molecular Dynamics Supercomputer, D. E. Shaw et al., Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC14), (2014)

[2] Accurate and Efficient Integration for Molecular Dynamics Simulations at Constant Temperature and Pressure, R. A. Lippert et al., Journal of Chemical Physics, vol. 139, no. 16, 2013, pp. 164106:1–11.

[3] Desmond/GPU Performance as of October 2015, M. Bergdorf et al., D. E. Shaw Research Technical Report DESRES/TR—2015-01, 2015 (available at http://www.deshawresearch.com/publications.html)

[4] http://www.deshawresearch.com/publications.html