Grid Computing for Long Time Propagations in Molecular Dynamics of Biomolecules and Extended Phase Space Sampling

Scientists from Department of Chemistry, University of Crete (UoC) and 
Institute of Electronic Structure and Laser, 
FOundation for Research and Technology Hellas (FORTH)  have used the HellasGrid Infrastructure and the EGI Grid infrastructure in order to solve problems coming from the area of Computational Chemistry.

More specific, the scientists made studies on dynamics and spectroscopy of proteins and other biomolecules, enzymatic reactions, and free energy hypersurface calculations for protein – ligand interactions by integrating very long time trajectories and making extended sampling of the classical phase space [1-3].

For problems which exhibit ergodicity in a submanifold of the phase space manifold of the system, time averages are converted to phase space averages, and thus to high throughput problems. The algorithm that have been developed is based on the principle to run short jobs, store intermediate results, and resubmit the jobs as many times as needed to achieve a predetermined total integration time [1]. This approach cures shortcomings of the current productive Grid.

The production of a free energy hypersurface, requires one to run thousands or even millions of jobs, a task which can be fulfilled if thousands of CPUs are available in a reasonable amount of time [2,3]. Scripts based on the gLite middleware have been developed and tested at the HellasGrid and EGI infrastructure which can handle such tasks.

Quantum molecular dynamics for intermediate size biomolecules require a interoperability of high throughput and high performance computing. This is scientists’ plan for solving problems involving proton-coupled electron transfer (PCET), which are encountered in enzyme reactions, fuel cells, chemical sensors and electrochemical devices.

Contact details

Members of the Theoretical and Computational Chemistry group in Crete (TCCC):

  • Prof. Stavros C. Farantos, Theoretical and Computational Chemistry, farantos (at) iesl.forth.gr
  • Mr. Manos Giatromanolakis, IT Manager and Systems Analyst, gmanos (at) iesl.forth.gr
  • Prof. Vangelis Daskalakis, Computational Biochemistry, chem487 (at) edu.uoc.gr
  • Dr. Osvaldo Gervasi, Computer scientist, administrator of Compchem Virtual Organization,  osvaldo (at) unipg.it
  • Dr. Massimiliano Porrini, Theoretical and Computational Chemistry, maxp (at) iesl.forth.gr
  • Dr. Jaime Suarez, Theoretical and Computational Chemistry, jaime.suarez (at) iesl.forth.gr

References

  1. V. Daskalakis, M. Giatromanolakis, M. Porrini, S. C. Farantos, and O. Gervasi, Computer Physics, Chapter: Grid computing multiple shooting algorithms for extended phase space sampling and long time propagation in Molecular Dynamics, Nova Science Publishing Co., 2011.
  2. M. Porrini, V. Daskalakis, and S. C. Farantos, Thermodynamic Perturbation Calculations on Cytochrome c Oxidases interacting with small ligands, Phys. Chem. Chem. Phys., submitted, 2011.
  3. Massimiliano Porrini, Vangelis Daskalakis, S. C. Farantos, and Constantinos Varotsis, Heme Cavity Dynamics of Photodissociated CO from ba3-Cytochrome c Oxidase: the Role of Ring-D Propionate, J. Phys. Chem. B, 113(35):12129-12135, 2009.

 

 

Reinforcement Learning Game – RLGame


Scientists from Hellenic Open University and Computer Technology Institute and Press “Diophantus” have used the HellasGrid Infrastructure and the EGI Grid infrastructure in order to solve problems coming from the area of Machine Learning and Algorithms in Games.

When using machine learning to learn how to play a game, vast amounts of experiments may be required to adequately explore the space of alternative tactics and strategies. To cover this need, grid was a technology which was exploited by the researchers.

The scientists have identified a curious phenomenon in game playing, which they term “the pendulum effect”. Therein, they use a tutor to provide playing advice to one player (A) of a zero-sum two-player game. It seems that the player (B) who does not receive tutoring is also able to enormously benefit just by being forced to play against a better player (A, as advised by the tutor). This is reinforced when the tutor abstains; player B seems to be able to win more often as compared to when the tutor is present directing A to more wins.

By using the HellasGrid and EGI infrastructure, the scientists conducted a range of experiments across hundreds of game configurations and across several learning schemes; some of them employed game tree look-ahead search which is pretty expensive (but not yet optimized for grid computing). These experiments may have consumed several tens of thousands of CPU hours on the grid; such availability of resources simply does not exist at stand-alone venues.

Future plans include the continuation of this line of research to verify this type of learning behavior, across more configurations and actions to verify it across other games as well. The scientists are also currently looking into workflows (hopefully, grid-aware workflow systems) to better organize data collection and analysis.

Contacts

  • Dimitris Kalles, Hellenic Open University, kalles (at) eap.gr
  • Panagiotis Kanellopoulos, Computer Technology Institute and Press “Diophantus”, kanellop (at) ceid.upatras.gr
References
  1. D. Kalles, and I. Fykouras. “Examples as Interaction: On Humans Teaching a Computer to Play a Game”, International Journal on Artificial Intelligence Tools, 2010.
  2. D. Kalles and P. Kanellopoulos. “A Minimax Tutor for Learning to Play a Board Game”, Workshop on Artificial Intelligence in Games, a workshop of the 18th European Conference on Artificial Intelligence, 2008.
  3. D. Kalles and P. Kanellopoulos. “A Pendulum Effect in Co-evolutionary Learning in Games”, European Workshop in Reinforcement Learning, 2011.

 

Users Support Team

The main purpose of HellasGrid User Support Team (UST) is to guide the new users on the procedure of accessing the HellasGrid infrastructure and their continuous support during their daily work with the grid infrastructure. For a candidate user to access the HellasGrid infrastructure, he/she must register at the infrastructure at the site https://access.hellasgrid.gr. Detailed guides concerning the access procedure are available here.

Specifically, the HellasGrid Users Support Team provides the following services in collaboration with other support groups of HellasGrid:

  • Consulting services: The UST provides consulting services to users wishing to access the grid infrastructure of HellasGrid.
  • User Support: UST is a first level of support for the users to solve their problems concerning their daily work with grid infrastructure and serve their requests, in general.
  • Users’ request gathering. UST performs periodically various surveys for gathering users’ requirements and their satisfaction degree by the HellasGrid infrastructure. The results of these surveys are used as the base for the further development of the grid infrastructure and the improvement of the provided services.
  • Creation of supportive users’ guides. UST in collaboration with the other support teams of HellasGrid creates various guides and documentation pages concerning users’ support.
  • Provision of statistics. UST provide statistical reports concerning the registered to HellasGrid users, for example their number, their scientific fields, their institute/organization, etc.

Furthermore, other support tools have been created:

  • HellasGrid wiki: It is the official HellasGrid wiki, which the users can use in order to find supportive information concerning the access to HellasGrid infrastrucute and its use.
  • Software discovery engine: It is a web portal from where the users can search at which HellasGrid sites and User Interfaces specific software packages are installed .

The users can come in contact with HellasGrid UST by using the following e-mail user-support (at) hellasgrid.gr.