Hello,
How do you all normally analyze multiple MD simulation trajectories. For my system, I ran 5 100 ns simulations using an different initial velocity. Do you perform cluster analysis and then perform RMSD, RMSF etc. from the most popular state or do you analyze each trajectory separately and then average the results?
5 trjs are not enough for clustering - but maybe you can check how much they converged - and then measure the mismatch between your expectation before and after seeing the trjs. This helps measure the uncertainty of the simulations.
Ramin Ekhteiari Salmas I am looking to do something similar to the picture attached. Plotting the average RMSD, RMSF, etc. among the 5 replicates and showing the standard error with the shaded region. I believe these were performed using MDTraj. Does anyone have sample scripts for performing something similar and can this also be done in cpptraj?
tsplot:
https://www.adamsmith.haus/python/docs/seaborn.tsplot
- Badges
- Science topic