ISyMAB

From HP-SEE Wiki

Revision as of 09:55, 1 April 2012

General Information

• Application's name : Integrated System for Modeling and data Analysis of complex Biomolecules
• Application's acronym: ISyMAB
• Virtual Research Communities : Computational Chemistry
• Scientific contact: Ionut Traian Vasile, itvasile@nipne.ro
• Technical contact: Ionut Traian Vasile, itvasile@nipne.ro
• Developers: Dragos Ciobanu-Zabet, Ionut Traian Vasile, Department of Computational Physics and Information Technologies, IFIN-HH, Romania
• Web site: http://wiki.hp-see.eu/index.php/ISyMAB

Short Description

We aim at providing an efficient tool for data analysis in the framework of the simulation and modeling of large biomolecular systems by means of parallel molecular dynamics codes designed for high-performance computing (NAMD). The main focus is in the modeling of the G Protein-Coupled Receptors, which are found in a wide range of species, and are involved in signalling from the outside to the inside of the cells.

Generally, large biomolecule simulations with NAMD involve many steps which are not always easy and/or within the reach of the researchers. The proposed application provides an integrated solution that will help them solve these problems instantly.

The application builds a remote access framework on NAMD clusters which offers the users an integrated interface with analysis tools, which will make possible fast data analysis of complex simulation outputs directly on the remote systems. The users provide the initial parameters of the MD simulation, launch jobs, and make various dataanalysis on the output files.

The application will generate the inputs needed for the batch system, the inputs for the MD simulation, will help manage simulation jobs, make various data analysis on the outputs obtained, visualize the biosystem after the job, and make graphs on the selected data. Besides this, the application will ensure the secure connection of the user with the cluster.

Problems Solved

The application facilitates the interaction between the scientists and the NAMD cluster, making the simulation work on clusters much easier and user-friendly. Also, the system will provide and integrate an extensive list of publicly available tools that are used in many research centers which simulate large biomolecules using NAMD.

Scientific and Social Impact

The integrated system will increase the user productivity, defined as the ratio between the actual simulation time and the total time spent for running the simulation. Also, it will accelerate research, the scientists getting relevant results sooner.

The application will help the workers/students in biophysics and biochemistry to perform faster research and discovery, by optimizing the time needed for MD simulations.

Collaborations

• The integrated system will be used in the research collaborations started with the Faculty of Physics and the Faculty of Biology of the University of Bucharest.

Beneficiaries

• Reasearchers in the field of biophysics and biochemistry.

Number of users

2

Development Plan

• Concept: Done before the project started.
• Start of alpha stage: Done before the project started.
• Start of beta stage: M12
• Start of testing stage: M17
• Start of deployment stage: M19
• Start of production stage: M21

Resource Requirements

• Full-scale number of logical CPUs : 256
• Minimum RAM/core required : 512 MB
• Storage space during a single run : 10 GB
• Long-term data storage : 10 TB
• Total core-hours required: 300.000

Technical features and HP-SEE implementation

• Primary programming language : C
• Parallel programming paradigm : asynchronous message passing (MPI/CHARM++)
• Main parallel code : NAMD 2.7b3 http://www.ks.uiuc.edu/Research/namd/
• Pre/post processing code : In-house developed scripts plus existing scripts/software like:

    - MMTSB tool set, http://blue11.bch.msu.edu/mmtsb
    - VMD 1.8.7, http://www.ks.uiuc.edu/Research/vmd/

Usage Example

• login into the application's web interface at https://hpc.ifin.ro/
• insert username and password or use the SSL certificate
• create files using the provided templates (for a PBS job, NAMD input or a custom shell script)
• manage files located locally on the webserver, on the Permanent Storage or on the Internal Cluster Storage; here you can Edit, Delete, Download, Upload, Copy, or send the file to Execute on the cluster
• view the running jobs on the PBS job manager and the details for each queue and each job; also you can check the general informations for each PBS server and each queue
• check settings for NFS file sharing and VNC server
• use MMTSB (Multiscale Modeling Tools for Structural Biology) module
• use download module for downloading PDB files directly from the Protein data bank (http://www.rcsb.org)
• check statistical graphs

Infrastructure Usage

• Home system: IFIN_Bio
  • Applied for access on: .
  • Access granted on: .
  • Achieved scalability: .
• Accessed production systems:

1. IFIN_BC
   • Applied for access on: .
   • Access granted on: .
   • Achieved scalability: .

• Porting activities: The application has been successfully ported from the IFIN_Bio cluster system to the IFIN_BC cluster.
• Scalability studies: Tests up to 256 cores, on IFIN_Bio.

Running on Several HP-SEE Centres

Achieved Results

Publications and Presentations

• Ionut Vasile, Dragos Ciobanu-Zabet, Software tools for HPC users at IFIN-HH, RO-LCG 2011 Workshop Applications of Grid Technology and High Performance Computing in Advanced Research, November 29-30, 2011, http://rolcg11.nipne.ro/presentations/ROLCG_2011-software_tools_and_Isymab.pdf

• Drakulić Branko J., Dragos Ciobanu-Zabet, Ionut Vasile, Ivanov Petko, Dodoff Nicolay, Karaivanova Aneta, Juranić Ivan O., "The Medicinal Chemistry-related applications in HP-SEE project. An End-user view", Hellenic Symposium on Medicinal Chemistry, Athens, May 25-27, 2012 http://www.helmedchem2012.gr

Foreseen Activities

• Testing of the integrated IFIN_Bio + IFIN_BC interface in view of the production stage.