Anton, named for Anton van Leeuwenhoek, is a special purpose supercomputer for molecular dynamics (MD) simulations, designed and constructed by D. E. Shaw Research (DESRES). In colllaboration with DESRES, the Biomedical Applications Group at PSC is hosting an Anton machine for general availability to the national biomedical community.
Anton can carry out MD simulations at an atomic level and on the scale of milliseconds - and at speeds up to 100 times faster than a conventional supercomputer.
The availability of Anton at PSC was made possible through the generosity of D.E. Shaw and grants from the NIH's National Institute of General Medical Sciences.
Access to Anton
Anton is allocated annually with proposals reviewed by a committee convened by the National Research Council at the National Academies. 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.
The next request for proposals (RFP) for time on Anton is expected in late spring 2013 and will be announced on the PSC web site.
Anton uses specialized hardware to perform molecular dynamics simulations orders of magnitude faster than general purpose hardware running traditional MD software.
Anton is designed primarily to accelerate classical MD simulations of biomolecular systems with periodic boundary conditions and explicit solvent. Projects that maximize the benefit of Anton to the scientific community focus on questions that will be greatly advanced by multi-microsecond MD simulations, and especially those questions that require long continuous trajectories rather than a sampling of many short ones.
Simulations are standard MD runs in the NVE, NVT (isothermal), and NPT (isothermal, isobaric) ensembles. They may use Berendsen or Nose-Hoover thermostats, and may use Berendsen or MTK barostats with isotropic or semi-isotropic scaling. 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 a restraint between the centers of mass of groups of atoms, and (iii) application of a restraint based on the calculation of RMSD (root mean squared deviation) to atomic positions of a reference structure. The total number of atoms involved in either distance restraints or RMSD restraints may not exceed 2048.
Typically, 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. In most cases, the ratio of the largest side to smallest side must not exceed 1.5:1.
Simulations use recent variants of the following standard biomolecular force fields: CHARMM (e.g., CHARMM22, CHARMM27 - including CMAP corrections, and CHARMM36 for lipids), 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 is modeled with the SPC, TIP3P, or TIP4P models, or their variants.
Chemical systems should contain between 20,000 and 120,000 atoms (including solvent atoms), and typically consist of some combination of protein, DNA, RNA, lipids, water, and standard ions.
Anton enables scientists to perform MD simulations of biomolecular systems nearly two orders of magnitude faster than the previous state of the art. This table shows benchmark information for a number of systems of various sizes. Performance was measured on a 512-node Anton machine like the one hosted by the PSC. Note that Anton is a single user system, i.e., simulations on Anton use all 512 nodes.
|Chemical system (PDB ID)||Number of atoms||Approximate performance (microseconds/machine-day)*|
* All simulations used 2.5-femtosecond time steps with long-range interactions evaluated at every other time step and a Berendsen thermostat applied every 100 time steps.
Research conducted on Anton
Listed here is just some of the groundbreaking research enabled by Anton:
- Millisecond-scale molecular dynamics simulations on Anton - This paper (Gordon Bell prize winner for best paper at SC09 ) contains measurements of energy conservation on Anton that you can use to compare with your own simulations.
- Atomic-Level Characterization of the Structural Dynamics of Proteins - This paper, published in Science, details the first millisecond MD simulation on Anton.
- Protein Research Leaps Forward - four projects in MD simulation from PSC's 2011 annual report, Projects in Scientific Computing
- Epic Microseconds - four projects yielding invaluable insights into the structure and function of proteins from PSC's 2012 annual report, Projects in Scientific Computing
Here is a list of publications for research that made use of Anton at PSC:
- Long time-scale Molecular Dynamics simulations elucidate the dynamics and kinetics of exposure of the hydrophobic patch in Troponin C., Lindert, S., Kekenes-Huskey, P.M., and McCammon, J.A. (2012). Biophys J, In press.
- Microscopic Origin of Gating Current Fluctuations in a Potassium Channel Voltage Sensor, J. Alfredo Freites, Eric V. Schow, Stephen H. White, and Douglas J. Tobias, 2012, Biophys J 102, L44–L46.
- Structural Characterization of λ-Repressor Folding from All-Atom Molecular Dynamics Simulations Yanxin Liu, Johan Strümpfer, Peter L Freddolino, Martin Gruebele, and Klaus Schulten, 2012 J. Phys. Chem. Lett., 3, 1117–1123.
- Molecular dynamics study of MspA arginine-mutants predicts slow DNA translocations and ion current blockades indicative of DNA sequence, Swati Bhattacharya, Ian M. Derrington, Mikhail Pavlenok, Michael Niederweis, Jens H. Gundlach, and Aleksei Aksimentiev, 2012, ACS Nano, 6:6960–6968.
- Long-timescale dynamics and the regulation of Sec-facilitated protein translocation, B. Zhang and T. F. Miller, III, Cell Rep., in press.
- Direct simulation of early-stage Sec-facilitated protein translocation, B. Zhang and T. F. Miller, III, 2012 J. Am. Chem. Soc. 134, 13700-13707.
- Membrane-binding Mechanism of a Peripheral Membrane Protein through Microsecond Molecular Dynamics Simulations, B Rogaski and J. B. Klauda, 2012, J. Mol Biol, in press.
- Structure, orientation, and surface interactions of Alzheimer amyloid-β peptides on the graphite, X. Yu, Q. Wang, Y. Lin, J. Zhao, C. Zhao, and J. Zheng, 2012, Langmuir 28: 6595-6605.
- Probing ion channel activity of human islet amyloid polypeptide (amylin), J. Zhao, Y. Luo, H. Jang, X. Yu, G. Wei, R. Nussinov, and J. Zheng, 2012, BBA Biomembranes 12: 3121-3130.
- Mechanism of Drug Efficacy within the Epidermal Growth Factor Receptor Revealed by Microsecond Molecular Dynamics Simulation, S. Wan, D. Wright, P. V. Coveney, 2012, Molecular Cancer Therapeutics, Available Online, DOI: 10.1158/1535-7163.MCT-12-0644-T.
- From base pair to bedside: molecular simulation and the translation of genomics to personalised medicine, D. W. Wright, S. Wan, N. Shublaq, S. Zasada, P. V. Coveney, 2012, WIREs Syst Biol Med. doi: 10.1002/wsbm.1186.
- Learning Generative Models of Molecular Dynamics. N. S. Razavian, H. Kamisetty, C.J. Langmead, 2012 BMC Genomics 13: S1-S5.
- Mechanistic Insights into αIIbβ3 Integrin Signaling from Long-Timescale Molecular Dynamics Simulations", Provasi, D., Negri, A., Coller, B.S., Filizola, M. (2012) Biophys J 102, 398a - 399a.
- Microsecond Dynamics of the G-Protein Coupled Receptor Squid Rhodopsin in Atomistic Detail, M. Heyden, E. Jardón-Valadez, A. N. Bondar, D.J. Tobias, (2012) Biophys J 102; 622a.
- Structure of a force-conveying cadherin bond essential for inner-ear mechanotransduction, Marcos Sotomayor, Wilhelm A. Weihofen, Rachelle Gaudet, David P. Corey, (2012), Nature, 492, 128–132.