Report of the Committee on Proposal Evaluation for Allocation of Supercomputing Time for the Study of Molecular Dynamics: Fourth Round (2013)

September 18, 2013

Jodi Swidzinski Hezky, Ph.D.
D. E. Shaw Research
120 West 45th Street, 39th Floor
New York, NY 10036

Dear Dr. Hezky:

This letter describes the work and transmits the final report of the Committee on Proposal Evaluation for Allocation of Supercomputing Time for the Study of Molecular Dynamics, Fourth Round.

The committee evaluated submissions received in response to a Request for Proposals (RFP) for Biomolecular Simulation Time on Anton, a supercomputer specially designed and built by D.E. Shaw Research (DESRES) that allows for dramatically increased molecular dynamics simulations compared to other currently available resources. Over the past three years (October 1, 2010 – September 30, 2013), DESRES has made available to the non-commercial research community node-hours on an Anton system housed at the Pittsburgh Supercomputing Center (PSC), based on the advice of previous National Research Council committees convened in 2010, 2011, and 2012.

The success of the program has led DESRES to make the Anton machine housed at PSC available for an additional 3,700,000 node-hours over the period following October 2013, and DESRES has asked the National Research Council to once again facilitate the allocation of time to the non-commercial research community. The work of the National Research Council committee to evaluate proposals for time allocations was supported by a contract between D.E. Shaw Research and the National Academy of Sciences and was performed under the auspices of the National Research Council’s Board on Life Sciences.

To undertake this task, the National Research Council convened a committee of experts to evaluate the proposals submitted in response to the RFP. The committee of 15 was chaired by Dr. Michael Levitt (NAS), Professor and Chair, Department of Structural Biology, Stanford University School of Medicine. The committee members were selected for their expertise in molecular dynamics simulations and their experience in the subject areas represented in the 65 proposals that were considered by the committee. They comprised a cross section of the biomolecular dynamics field in academia, industry, and government including an array of both senior and junior investigators.

The goal of the fourth RFP for Biomolecular Simulation Time on Anton has been to continue to facilitate breakthrough research in the study of biomolecular systems by providing a massively parallel system specially designed for molecular dynamics simulations. These special capabilities allow multi-microsecond simulation timescales, which previously had been unobtainable. The program seeks to continue to support research that addresses important and high impact questions demonstrating a clear need for Anton’s special capabilities.

The Anton RFP described the three criteria against which the committee was asked to evaluate proposals:

Proposals from investigators who had previously received an allocation of time on Anton were required to include progress reports. Following guidance provided by DESRES and PSC, the committee drew on these progress reports as supplemental material in its consideration of proposals. The committee also received information from PSC on the number of node-hours of simulation time remaining on 2012 Anton allocations (as of July 17, 2013), as further supplemental information. As explained in the RFP, staff at PSC conducted an initial assessment of all proposal submissions for completeness and to determine whether they were technically feasible for simulation on Anton. A member of the PSC staff was present as an observer throughout the review committee’s discussions to address any additional questions that arose on Anton’s technical capabilities or on how the computer will be made available to researchers during the period of the project.

The committee was asked to identify proposals that best met the selection criteria defined above. As in the previous rounds of Anton time allocations, 100,000 node-hours was the maximum amount of time available to a proposal. Principal investigators could also request a lesser time allocation. The committee was further asked to allocate at least 25% of the time to principal investigators who had not previously received an Anton allocation. The judgments of the committee are based on which proposals best met the selection criteria described above and on the estimates of required simulation time provided by the applicants. The committee was permitted to consider a modified time allocation if it concluded that the proposed research required a greater or lesser number of node-hours than initially requested by an applicant.

Initial reviews of the proposals were provided by the 15 committee members. Each proposal was assigned a minimum of two primary reviewers who were asked to evaluate the proposal based on the RFP and guidelines described above. Review assignments were made so that no proposal was

evaluated by a reviewer from the applicant’s same institution or who had a collaborative relationship with an applicant.

The committee held its meeting in Washington, D.C. on August 5, 2013. At the meeting, members undertook a detailed discussion of the proposals. The two primary reviewers were asked to summarize their review for the committee, which was followed by discussion of the proposed research. As described in detail above, committee members considered the scientific merit, justification of the requested time, and the qualifications of the principal investigator. The committee then considered the slate of proposals, came to a consensus on which proposals it judged best met the selection criteria, and, in some cases, decided to suggest a modified allocation of time on Anton. Detailed comments for each of the 65 proposals are included in Appendix B.

The committee concluded that the proposals listed below best met the selection criteria set forth in the RFP for Biomolecular Simulation Time on Anton. Of these 48 proposals, 15 proposals were selected for a modified allocation (identified below with an *). For investigators who received an Anton allocation in 2012 that had over 50% of the simulation time remaining (based on the data provided to the committee by Pittsburgh Supercomputing Center), the committee automatically reduced the 2013 allocation request by 10%.

In numerical order by proposal submission number, the proposals judged by the committee as best meeting the selection criteria of the RFP are:

PSCA13003P The calcium dependent regulation of thin filament protein troponin – molecular dynamics insights of troponin relaxation and activation pathways; PI: Lu, University of Illinois at Chicago [New user, identified for 25,000 node-hours]^*

PSCA13005P Microsecond scale simulations to characterize skeletal muscle Ca2+-binding proteins; PI: McCammon, University of California San Diego [Returning user, identified for 45,000 node-hours]^*

PSCA13006P Exploring Roles of Glycans in Interactions between HIV-1 gp120 and Broadly Neutralizing Antibodies; PI: Im, University of Kansas [Returning user, identified for 100,000 node-hours]

PSCA13007P Alternating access dynamics in lactose permease (LacY); PI: Kaback, University of California Los Angeles [New user, identified for 50,000 node-hours]^*

PSCA13008P Ensemble Modeling of Intrinsically Disordered Proteins: DNA Binding by Pdx1; PI: Showalter, Pennsylvania State University [Returning user, identified for 100,000 node-hours]

PSCA13009P Molecular Machinery Controlling the Activation and Clamping of Synaptic Vesicle Fusion; PI: Bykhovskaia, Universidad Central del Caribe [New user, identified for 100,000 node-hours]

PSCA13011P Investigating the atomic detail mechanisms of mutagenesis modulated cation selectivity of bacterial voltage-gated sodium channels; PI: Clapham, HHMI/Cardiology Boston Children's Hospital [New user, identified for 40,000 node-hours]^*

PSCA13012P Visualizing dynamic movements in heme proteins; PI: Poulos, University of California, Irvine [New user, identified for 50,000 node-hours]^*

PSCA13013P Molecular details of the activation of the μ opioid receptor; PI: MacKerell, University of Maryland School of Pharmacy [Returning user, identified for 100,000 node-hours]

PSCA13014P Atomistic mechanism for the microtubule-activated kinesin ATPase; PI: Hwang, Texas A&M University [Returning user, identified for 100,000 node-hours]

PSCA13015P Microsecond simulation studies of mechanochemical coupling in the molecular motor myosin; PI: Cui, University of Wisconsin-Madison [New user, identified for 99,533 node-hours]

PSCA13017P Calculating Conductance of Ion Channels; PI: Pohorille, University of California San Francisco [Returning user, identified for 50,000 node-hours]

PSCA13018P Using Microsecond Simulations to Characterize Cholesterol Interactions with Chemokine Receptors; PI: Handel, University of California San Diego [New user, identified for 50,000 node-hours]^*

PSCA13019P Peptide-induced Pore Formation in Lipid Membranes; PI: Lazaridis, City College of New York [New user, identified for 99,533 node-hours]

PSCA13023P Long Time Scale Molecular Dynamics Simulation of Protein Folding; PI: Gruebele, University of Illinois at Urbana-Champaign [Returning user, identified for 90,000 node-hours]^*

PSCA13024P Gold-Standard Conformational Transitions of Adenylate Kinase; PI: Beckstein, Arizona State University [New user, identified for 50,412 node-hours]

PSCA13025P The Correlated Dynamics of the Tandem Cyclic Nucleotide Binding Domains in the Regulatory Subunit of Protein Kinase A; PI: Amaro, University of California, San Diego [Returning user, identified for 100,000 node-hours]

PSCA13029P Dynamics of Bacterial Transcription Initiation; PI: Thirumalai, University of Maryland [Returning user, identified for 100,000 node-hours]

PSCA13031P Evolutionary Pathways of Engineered Sitagliptinases through Microsecond Molecular Dynamics; PI: Houk, University of California, Los Angeles [Returning user, identified for 100,000 node-hours]

PSCA13032P Determining the mechanisms of protein folding in membranes; PI: Gumbart, Georgia Institute of Technology [Returning user, identified for 100,000 node-hours]

PSCA13033P The molecular determinants of selective ion binding in the sodium-potassium pump ATPase; PI: Roux, University of Chicago [Returning user, identified for 90,000 node-hours]^*

PSCA13034P Resolving specific structural implications of functional mechanisms of the Leucine Transporter with long MD simulations; PI: Weinstein, Weill Cornell Medical College of Cornell University [New user, identified for 25,000 node-hours]^*

PSCA13035P Pathway and mechanism of drug binding to HIV-1 protease; PI: Chang, University of California, Riverside [New user, identified for 73,396 node-hours]

PSCA13036P Investigation of long time protein dynamics under physiological conditions; PI: Cheng, University of Tennessee, Knoxville [New user, identified for 100,000 node-hours]

PSCA13037P Simulations of the HER2-HER3 heterodimeric kinase complex: Understanding the activation mechanism and its role in cancer therapy; PI: Sept, University of Michigan [New user, identified for 100,000 node-hours]

PSCA13038P Dynamics of the Early Translational Machinery; PI: Luthey-Schulten, University of Illinois at Urbana-Champaign [Returning user, identified for 100,000 node-hours]

PSCA13039P Ligand-Binding Mechanics in a Glutamate Receptor; PI: Lau, Johns Hopkins University School of Medicine [New user, identified for 100,000 node-hours]

PSCA13040P A novel approach for simulating the complete transport cycle of neurotransmitter:sodium symporters; PI: Bahar, University of Pittsburgh School of Medicine [Returning user, identified for 100,000 node-hours]

PSCA13041P Anomalous transport and lateral structure in physiological, nanodomain forming lipid mixtures; PI: Lyman, University of Delaware [Returning user, identified for 100,000 node-hours]

PSCA13043P An unconventional approach to rhodopsin activation using detergent micelles; PI: Mertz, West Virginia University [New user, identified for 50,000 node-hours]

PSCA13044P Kinetic studies of heterogenous folding pathways in a hyper-stable RNA motif; PI: Garcia, Rensselaer Polytechnic Institute [New user, identified for 100,000 node-hours]

PSCA13046P Computing the Thermodynamics of HIV-1 gp41-Mediated Viral Membrane Poration using Microsecond MD Simulation; PI: Abrams, Drexel University [Returning user, identified for 35,000 node-hours]^*

PSCA13048P Simulations of a Peripheral Membrane Protein Binding Mechanism to Yeast Organelle Membranes and Forming Membrane Contact Sites; PI: Klauda, University of Maryland [Returning user, identified for 90,000 node-hours]^*

PSCA13050P Modeling the Antibody-Eliciting Conformation of the HIV gp41 MPER; PI: DeGrado, University of California San Francisco [New user, identified for 98,835 node-hours]

PSCA13051P 20 μs simulation of bc1 complex of mitochondrial respiratory chain; PI: Matyushov, Arizona State University [Returning user, identified for 55,000 node-hours]^*

PSCA13052P Molecular mechanisms of geometry-based intracellular organization; PI: Huang, Stanford University [New user, identified for 25,000 node-hours]^*

PSCA13053P Understanding the pathway of integrin headpiece opening; PI: Springer, Harvard Medical School and Children's Hospital Boston [Returning user, identified for 100,000 node-hours]

PSCA13056P Characterizing the disordered FG repeat domains of Nuclear Pore Complexes by simulation and experiment; PI: Cowburn, Albert Einstein College of Medicine, Yeshiva University [New user, identified for 100,000 node-hours]

PSCA13062P Membrane Insertion and the Equilibrium Configurational Ensemble of Histidine Kinase Receptors; PI: White, University of California, Irvine [Returning user, identified for 95,581 node-hours]

PSCA13063P Co-translational folding of nascent protein on the ribosome; PI: Schulten, University of Illinois at Urbana-Champaign [Returning user, identified for 100,000 node-hours]

PSCA13065P Detailed characterization of the structure and conformational dynamics of full-length HIV-1 gp120 in the unliganded form; PI: Langmead, Carnegie Mellon University [Returning user, identified for 50,000 node-hours]

PSCA13066P Molecular modeling studies of drug binding to human P-glycoprotein; PI: Freites, University of California-Irvine [Returning user, identified for 100,000 node-hours]

PSCA13068P An investigation into intramolecular diffusion of a disordered protein; PI: Stultz, Massachusetts Institute of Technology [New user, identified for 50,000 node-hours]

PSCA13069P Microsecond Dynamics of HIV-1 Stem Loop1 RNA for computation of NMR order parameters and comparison with HIV-1 TAR RNA; PI: Andricioaei, University Of California at Irvine [Returning user, identified for 50,000 node-hours]

PSCA13070P Dynamics of the Resting State of the Ci-VSP Voltage Sensor Under Applied Transmembrane Potentials; PI: Perozo, University of Chicago [New user, identified for 100,000 node-hours]

PSCA13072P Polycyclic Natural Products and Analogues through Computational Enzyme Engineering: Upstream Design for New Polyketide Building Blocks; PI: Pande, Stanford University [Returning user, identified for 45,000 node-hours]^*

PSCA13074P Sequencing DNA Using MspA; PI: Aksimentiev, University of Illinois at Urbana-Champaign [Returning user, identified for 90,000 node-hours]^*

PSCA13077P Characterizing Ion-coupled Structural Transitions in Secondary Active Membrane Transporters; PI: Tajkhorshid, University of Illinois at Urbana-Champaign [Returning user, identified for 25,000 node-hours]^*

The time allocations for the 48 proposals identified by the committee as best meeting the selection criteria for time allocations total approximately 3,698,000 node-hours. Of the 48 proposals identified, 30 were identified at the approximately 100,000 node-hour level and 18 at the 25,000-50,000 node-hour level.¹ A total of approximately 1,488,000 node-hours were allocated to 21 proposals whose principal investigator did not receive time on Anton during the past three years (identified as “new users”). Approximately 40% of the available simulation time thus was allocated to new users of Anton. The remaining 2,211,000 node-hours are allocated to 27 proposals from investigators who had received allocations of time on Anton in previous rounds (identified as “returning users”).

In carrying out its task, the committee identified as many promising proposals as possible given the constraints on the total available simulation time.

___________________

¹ The 100,000 node-hour level is defined as proposals that were identified for 70,000 node-hours or greater. The 50,000 node-hour level is defined as proposals that were identified for less than 70,000 node-hours.

The committee would like to thank D.E. Shaw Research, the Pittsburgh Supercomputing Center, and all of the 2013 Anton applicants for the opportunity to assist in identifying the proposals best meeting the selection criteria for time allocations on the Anton machine. The committee members were universally enthusiastic about the potential advances in the field that are facilitated by Anton and are looking forward to seeing the important new results from the Anton users.

Sincerely,

Michael Levitt
Chair