Running Molecular Dynamics on Alliance clusters with AMBER: Lesson Design

Help Wanted

We are filling in the exercises below in order to make the lesson plan more concrete. Contributions (both in the form of pull requests with filled-in exercises, and comments on specific exercises, ordering, and timings) are greatly appreciated.

Motivation for the BST-Workshop series

The motivation behind this 3-part workshop/tutorial series is to give new grad students an introduction into preparing, running and analyzing MD simulations. The series is not meant to be a replacement for university courses in thermodynamics and statistical thermodynamics. Instead this series takes an applied approach and touches on theory only as much it is necessary to aid choosing reasonable simulation parameters.

The first lesson is a hands-on example on how to use GROMACS to prepare and submit a MD simulation. There are numerous good GROMACS tutorials available and maybe one of them could be adapted for this. In addition to general instructions and commands, there should be specific instructions on how to do this on Compute Canada clusters. As in Software Carpentry lessons, certain steps should be implemented as exercises for the learner - the correct answer can be hidden in an answer box.
Theory should be skipped over as much as possible and should be covered in the second module.
This lesson could be adapted for other MD-codes (e.g. NAMD) as well.

The second lesson is a small theory review to remind the learners of important concepts and how they influence the choice of simulation parameters. The aim is that people can avoid pit-falls and misunderstandings that are commonly made by novice MD users. Topics to cover are Periodic Boundary Conditions (and why there is no “outside” with a periodic box), thermostats/barostats, cut-offs, etc.
This should be no substitute for a formal course in statistical thermodynamics but should help users make more informed choices of simulation settings/parameters.

The third lesson is again a hands-on tutorial to use Python and the Python packages MDAnalysis, MDtraj, and NGLview to write their own analysis tools. While, for example GROMACS, comes with a very large number of analysis tools out-of-the-box, users often limit themselves to those tools and the options and variations that they offer. The above frameworks make it very easy to read MD-trajectories in different formats and get access to the coordinates and come up with fully customized analysis methods.

The mentioned lessons one and three are currently put on hold and are subject to be created at a later time. In the meantime we can recommend to work through online tutorials that are already available such as:

This module here will implement the second lesson focusing on theory and giving guidance on choosing good parameters for MD-Simulations.

Process Used

Stage 1: Assumptions

Stage 2: Desired Results

Topics to cover

(in no particular order; might need to be pruned)

Questions

How do I…

Skills

I can…

Concepts

I know…

Stage 3: Learning Plan

Summative Assessment

Lesson 1: