Initial notes

• Create a new word file to answer the questions in this lab and name it [LastNames]MD2.docx
• Download your MDT folder from our previous lab (model building)
  • Remember - we'll be referring to the MDT folder inside it as$MDT
  • RIGHT BEFORE YOUR GROUP LEAVES TODAY, make sure to save this folder to someone's pitt.box.com!!!
• Today we'll be using the Mac Terminal application for our commands
• For more information about any of the NAMD parameters we use ==> see the NAMD manual

Pre-lab fix

Open your $MDT/production/DYN00.namd file in TextWrangler and edit the very last line from run 5000000 to run 15000000

Minimization

Before starting the simulation, we'll run a procedure that attempts to minimize the potential energy of our system to the lowest possible point.. The minimization should solve any steric problems (such as overlapping atoms) that we inadvertantly introduced when building our model. If not resolved, these types of steric problems might cause our simulation to crash (the numerical integrators would become unstable).

Open up the NAMD configuration $MDT/minimization/MIN00.namd using Text Wrangler or your favorite text editor (but not Word or TextEdit). This file specifies a number of parameters NAMD will use when performing the minimization of our model. The only changes that you will need to make to this file is to specify the location of the coordinates and structure parameters (these are the names of the final coordinate {.pdb} and protein structure {.psf} files from the previous lab). Notice where these files are in relation to MIN00.namd, and replace the XXXX with the paths to the files. (hint - each path will begin with ../ )

After you have made the necessary modifications

1. save MIN00.namd
2. Open Terminal, and use cd to change to $MDT/minimization
3. Run the minimization via:

1

namd2 +p2 MIN00.namd > min00.log

• If you don't get any response/error - that's a good sign!
• The minimization should take about 10min to complete
• Once it does complete (when your prompt appears again), a couple new files should appear in your minimization directory: min.log & min00.dcd.

Minimization Analysis

First we'll check the .log file it created to ensure that the gradient of the energy is less than 5.0 by the end of the minimization. Although it's just a text file, it is much easier to extract the pertinent lines using the unix command grep:

1

grep GRADIENT min00.log

Did the gradient of the energy get to less than 5.0 by the end of the minimization?

[SKIP the remainder of the minimization analysis for now - get your Heating command below started first]

We can visualize this change in the energy of the system using VMD:

1. In the VMD Main window, Extensions > Analysis > NAMD Plot
2. Use the file pull down menu to select your NAMD log file: min.log
3. We want to consider the total energy, as well as the contributions from the bond, angle, dihedral, electrostatic and vdw terms. (check each of these boxes)
4. Plot the data using the File pull down menu.

QUESTION 4: Does a single energy term dominate the total change in energy of the system? If so, which one?

We can also visualize our minimization using the min00.dcd file:

1. In the VMD Main window, File > New Molecule (Load your coordinate - pdb -file used as input for our minimization)
2. Then File > Load Data Into Molecule... to load the .dcd file.
3. Notice at the very bottom of the VMD Main window are controls for flipping through the 500 steps of our minimization.

Heating

The next step is to thermalize the system (i.e gradually heat it to room temperature - 300K), then gradually release the restraints we placed on the protein and the ion in the channel. We will accomplish this in three steps.

Take a quick look at $MDT/heating/HEAT00.namd in a text editor. Notice we're using the same structure and coordinate files ==> but are Continuing a job from the restart file../minimization/min00. In other words, we're starting from the min00.dcd created from our minimization step.

In the Terminal, change to the $MDT/heating directory, and run the first heating step:.

1

namd2 +p2 HEAT00.namd > heat00.log

Should take about 10min, and raise the temperature of our system to 25K.

The second heating step would take about an hour on these iMacs, so we're going to need some more power - in other words, our good friend Frank! So:

1. Connect to Frank using the Fugu or Fetch or WinSCP, and transfer your entire MDT directory over to your directory
• (if at home, or on Pitt-Wireless - see the VPN_to_Frank doc on our box share)
2. Connect to Frank using the Terminal (SSH)
3. Change into your $MDT/heating directory (on Frank)

If you remember, to run large jobs on Frank we'll need to get in line (ie: submit our job to the "queue"). So you'll first need to copy over a script from Dr. Adelman's account (that can submit your next heating step to the queue). The second line below will submit your job to the queue:

1
2

cp ~jla65/gramicidin/gram/heating/run_heat01.sh run_heat01.sh
  qsub run_heat01.sh

You can check your place in the queue by running the command: showq . On Frank this step should take about 6 minutes and raise the temperature of our system to 300K.

Do not go on to the next step until the following command says you have 0 active jobs and 0 eligible jobs: showq -w user=<username>

Our third and final heating step will keep the temperature constant, but slowly release the constraints we've placed on the system. As before, we'll first need to copy over a queue-ing script from Dr. Adelman's account - then run it:

1
2

cp ~jla65/gramicidin/gram/heating/run_heat02.sh run_heat02.sh
  qsub run_heat02.sh

On Frank, this step should take about 2 minutes ==> and now the system is ready for a lengthy simulation (often called the Production run)

[Skip the Heating Analysis step for now - we'll return to it after you have the main simulation started]

Heating Step(s) Analysis

1. Similarly to visualizing the minimization .dcd file above ==> load each heating step .dcd into VMD to visualize it.
2. Graphics ==> Representations ==> (Trajectory tab) ==> check Update Selection Every Frame . Then create a selection to visualize the water & potassium in the channel.
3. Use each log file with the NAMD Plot plugin to look at the temperature and volume of the system.
4. Save a copy of each plot using the Export to postscript option, and place in your <LastNames>Model directory.
5. Extensions > Analysis > RMSD Trajectory Tool to look at the root mean square deviation of the protein (see this link for details on using this tool). The RMSD is essentially a measure of how much your protein's atoms are moving at each step.

QUESTION 5: Qualitatively describe what is happening to the temperature,volume and rmsd during each simulation

Production

The system should now be ready for us to do a start a long simulation. This production run will take about 30 hours on Frank, and we'll analyze the resulting trajectories in our final lab session next Monday. So try to put it into the queue before leaving today.

In the Terminal, change to your$MDT/production directory on Frank, then:

1
2

cp ~jla65/gramicidin/gram/production/run_dyn00.sh run_dyn00.sh
  qsub run_dyn00.sh

HOMEWORK due before the start of the next class (Nov20):

• NOT YOUR ENTIRE folder, just your word document uploaded to our CompBio_2161_Drop