Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Sunhwan Jo Applicant's institution: ANL Applicant's division: LCF Project Name: entropy Project title: Quantifying protein-protein binding with greatly scalable multiple copy algorithms of NAMD Associated funding: None Other Systems: 0 (previously had 10M core hours in BG/Q, but is expired and submitted extension) Science: Protein-protein interactions (PPIs) represent a critical aspect of a wide range of biological processes, as most proteins need to associate to various binding partners to carry out their functions. While the basic principles of PPIs are generally understood, it is a grand challenge to accurately predict the binding affinity quantitatively for any given complex. Recently, this potential of mean force (PMF)-based methodology was extended to a much larger protein-peptide system, and we accurately determined the absolute binding free energy between the receptor protein and the ligand peptide. We plan to use our PMF-based methodology to probe effects of mutations in thermodynamics of protein-protein interaction, i.e., thermodynamic decomposition of barstar-barnase protein complex association. Project description: In PMF-based framework, a set of geometric and conformational restraints is applied to the proteins to reduce sampling noise and accelerate the convergence of the resulting free energy profile. These biases are rigorously accounted for to produce the absolute binding free energy. To enhance the PMF convergence, we will combine the umbrella sampling with Hamiltonian and temperature (ℋ and T) replica-exchange molecular dynamics (REM) algorithm. In the US/(ℋ,T)-REMD simulation approach, multiple copies of the molecular 
systems are simulated concurrently, but
 all with different
 conditions, e.g., umbrella potential (Hamiltonians) or temperature. Attempts are periodically made to exchange configurations between different systems using a Metropolis-type Monte Carlo acceptance criterion, insuring Boltzmann-weighted statistics. Such REMD
 methodology is extremely scalable, yet also very effective to sample a rugged free energy surface. We anticipate the simulation systems will be comprised of ~50,000 atoms and using at least 255 replicas concurrently. Because of its large size, we plan to use primarily Mira for production and Fusion will be used for initial system preparation and development of enhanced methodology in PMF-based framework. All planned simulations will be carried out by the open-source NAMD molecular dynamics software. The software engine used for the planned simulations is NAMD 2.10 with Charm++ 6.4.0. We have tested t he performance on BG/Q Mira. The scaling of parallel performance was linear up to 512 cores for our test system comprised of 100,000 atoms without using restraint potential and achieved the speed of about 0.025 s/step. Using restraint potential reduces the parallel performance significantly (~0.2 s/step). The expected number of project member is 3. I, Sunhwan Jo, recently joined LCF as a postdoctoral appointee and will be the main person performing code improvement and algorithm development under the supervision of Wei Jiang (LCF) and Benoit Roux (UChicago). Project URL: Requested allocation: 400000 Q1: 100000 Q2: 100000 Q3: 100000 Q4: 100000 Justification: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System