Atomistic Monte Carlo Simulations of Bio-molecular Systems
Jülich Supercomputing Centre, Forschungszentrum Jülich, Jülich, Germany
Organisers
Atomistic Markov Chain Monte Carlo simulations (MC) offer an interesting and complementary approach to Molecular Dynamics (MD) in the study of long time scale biophysical phenomena. MC simulations follow well established techniques to obtain a statistical physics description of a system using the potential energy function of the system, but without solving the equations of motion. While a temporal description of the simulated process is sacrificed, questions about probabilities, temperature dependence, effects of prevailing conditions etc. can often be answered at a minuscule fraction of the computing cost (and hence also the energy cost and carbon footprint) of an adequately long MD simulation. It has recently been demonstrated that replica exchange MC simulation can describe the folding of a complex 92 residue alpha + beta protein with an experimental folding time of about one second. We believe that MC techniques are severely underused in biophysical research, owing to the dearth of exposure to the technique in contrast to MD within the community. Our tutorial is an effort to introduce researchers in the field to MC techniques in sufficient detail for them to use it in their own research.
There is no fee for CECAM tutorials. Participants only need to cover their travel and accomodation. For a list of closeby hotels see Travel information and access to Jülich Supercomputing Centre.
Tutorial content
1. Theoretical background for Monte Carlo simulations
- Ergodicity, time scales, energy landscapes
- Monte Carlo fundamentals
- Detail balance
- Energy barriers and strategies to overcome them
- Comparison between MD and MC simulations
- Hybrid MC
2. Practical MC Simulation
- Introduction to ProFASi: the Protein Folding and Aggregation Simulator
- Folding a minimal protein with canonical MC simulations
3. Generalized ensemble simulations
- Simulated Tempering
- quality checks, statistics
- Replica Exchange MC simulations
- analysing REMC simulations (histogram reweighting, optimizing the
temperature ladder for replica exchange)
- Multi-canonical simulations
- The Wang-Landau method
4. Advanced introduction to ProFASi
- MC move sets
- Deformed energy landscapes (e.g. mechanical pulling)
- Non-protein molecules and non-standard residues in ProFASi
- Specifying interaction terms for non-standard residues and non-protein
molecules
- MC simulations with corrective restraint terms in the force field
- MC simulations with restraints in the conformational updates
- ProFASi’s plug-in interface: working with "dirty" experimental data
5. ProFASi as a library for experimentation with Monte Carlo techniques
- Defining your own force field in ProFASi
- How to introduce a new conformational update into ProFASi simulations
- How to try out new MC algorithms as ProFASi plugins
References
Jan Meinke (Forschungszentrum Jülich) - Organiser
Sandipan Mohanty (Forschungszentrum Jülich) - Organiser
Olav Zimmermann (Forschungszentrum Juelich) - Organiser