WESTPA2.0 with NAMD

The weighted ensemble (WE) strategy is an unbiased enhanced sampling technique that splits and reweighs trajectories. It is compiled in the Weighted Ensemble Simulation Toolkit with Parallelization and Analysis (WESPA). WESTPA2.0-NAMD is a modified version of the original WESTPA code by incorpoating NAMD for running simulations. WE_concepts.md includes my notes from Zuckerman and Chong, 2017.

Requirements

• WESTPA >= 2022.02
• NAMD >= 2.14
• Python >= 3.7

Getting started

After WESTPA2.0 is downloaded,

• common_files where you put config file, .psf, .pdb
• tstate file where you define target state
• bstate directory is where you put all ur restart .xsc, .vel, .coor
• west.cfg is where you define bins;
  • make sure pcoord lim equals the number of time you are outputting dcdfreq
  • bin_target_count equals number of starting trajectories

About Westpa_Scripts directory

• w_init calls for westpa_scripts/get_pcoord.sh so if there is an error edit that file.
• you need a pdb file of the last frame of your MD simulation so you can run get_pcoord.sh
• Then you submit jobs

Submitting Jobs

• NEW_sub.sh is submission script used in parallel with multiple GPU
• run.sh is submission script for running in serial
• w_run calls for westpa_scripts/runseg.sh. If you wanna stop or change anything you rename that and submit again.
• if you make error in anything delete the traj_seg of that iteration and the corresponding seg_log and you need to delete it from west.h5 using command w_truncate -n 100 if you want to get rid of anything after iteration #99.
• westpa_scripts/analysis.py can be used if w_pdist,w_succ, any w_ comands gives you an error about cannot read the h5 file.