De novo Peptide Folding in Membrane GB

Objective and Overview

This tutorial illustrates how to run implicit membrane GB with MMTSB toolset on ctbp cluster. The following is a step-by-step procedure to run replica exchange simulations in MMTSB toolset with implicit membrane GB model. If your job starts properly, you should have a new gbrex directory under walp16 directory. You may want to see what are in pept.inp and fold.sge files after you submit the job. Note that all necessary scripts and input files will be copied from the instructure's directory.

1. SGE Job Preparation and Submission

1. log onto ctbp cluster using
       ssh username@ctbp1.ucsd.edu
       
2. copy .cshrc from /home/jchen (if you already have one, append the contents to the existing .cshrc)
       cp /home/jchen/.cshrc .
       source .cshrc
       
3. copy necessary scripts and input files:
       cp -rp /home/jchen/walp16 .
       
4. go to walp16 directory
       cd walp16
       
5. generate an extended structure using CHARMM
       $CHARMMEXEC < scripts/pept.inp > pept.out
       
6. submit the job using qsub
	qsub scripts/fold.sge
	
7. you can check your job using ( in 'state' column, r means running and qw means on queue )
	qstat -u username
	

2. Result Analysis

In this tutorial, we will mainly focus on preliminary analysis using the MMTSB tools. If your job completed without any error, you should get 250 steps of replica exchange MD simulation.

1. How to look at the energy profile and which replicas are on a certain temperature using rexinfo.pl (try rexinfo.pl -help to see all available options).
rexinfo.pl -bycond 0 -dir gbrex

  Condition 0
  RUN  MODE RUN CLIENT TEMP    ENERGY       RMSD      VOLUME  
  1    PROD 1 aa1    300.0000  316.4390     8.0432    0.0000   
  2    PROD 2 aa1    300.0000  333.9761     7.2669    0.0000   
  3    PROD 3 aa1    300.0000  320.5962     6.4873    0.0000  
  4    PROD 4 aa1    300.0000  327.9228     6.2000    0.0000
  ....

  247  PROD 247  aa3    300.0000  259.3484     0.8795  0.0000   
  248  PROD 248  aa3    300.0000  259.8901     0.7706  0.0000   
  249  PROD 249  aa3    300.0000  241.5393     1.1392  0.0000   
  250  PROD 250  aa3    300.0000  247.1712     1.3147  0.0000
rexinfo.pl -bycond 0 -dir gbrex | tail +3 | awk '{print $1,$6}' > ens0.energy
rexinfo.pl -bycond 0 -dir gbrex | tail +3 | awk '{print $1,$7}' > ens0.rmsd
We can also compute the mean occupancies of clients,
rexinfo.pl -rankall -dir gbrex

  1  aa1 2.15 31.60
  2  aa2 3.24 2.80
  3  aa3 2.43 33.60
  4  aa4 2.18 32.00
...
One can also look at the energy and temperature profiles of each replica
rexinfo -byclient aa1 -dir gbrex

  Client aa1
  TRUN MODE RUN  COND TEMP      ENERGY       RMSD      VOLUME
  1    PROD 1    0    300.0000  316.4390     8.0432    0.0000
  2    PROD 2    0    300.0000  333.9761     7.2669    0.0000
  3    PROD 3    0    300.0000  320.5962     6.4873    0.0000
  4    PROD 4    0    300.0000  327.9228     6.2000    0.0000
  ....

  244  PROD 244  2    363.4241  316.8592     2.2206    0.0000   
  245  PROD 245  2    363.4241  288.2722     2.2581    0.0000   
  246  PROD 246  1    330.1927  288.0268     2.4723    0.0000   
  247  PROD 247  1    330.1927  306.1171     2.4880    0.0000   
  248  PROD 248  1    330.1927  268.5974     2.4799    0.0000  
  249  PROD 249  1    330.1927  294.0589     2.4003    0.0000  
  250  PROD 250  1    330.1927  267.8891     2.6742    0.0000
2. More trajectory analysis will be illustrated by additional tutorials on Friday.

Written by W. Im and J. Chen