search

dmzuckerman / Sampling-Uncertainty

Best Practices article intended for LiveCoMS
36 stars 5 forks source link

Equillibration and Quantification of Global Sampling #43

Closed dwsideriusNIST closed 6 years ago

dwsideriusNIST commented 6 years ago

Per reviewers:

"In section 5, the authors suggest to remove non-equilibrated trajectory regions from the beginning of simulations to improve simulation results. In my experience, this can indeed be very important and improve convergence of results a lot. Two references (27,28) for the determination of the time needed for equilibration are given, which both discuss algorithms for automated determination. I personally also find another, probably simpler approach proposed by Klimovich , Shirts & Mobley (Guidelines for the analysis of free energy calculations. J Comput Aided Mol Des. 2015) very useful. Simply speaking, in this approach forward and reverse cumulative averages are compared and data is discarded from the beginning of the trajectory until both match. With this approach, also visualization of convergence is nicely possible."

"Section 6.1 contains great information but I was left wanting more information and maybe a practical example to help solidify concepts."

"p11: "Relaxation should therefore…" -> I don't understand the sentence."

dwsideriusNIST commented 6 years ago

I'm happy with this revision, but would like another set of eyes to check it before closing the issue.