Use of Umbrella Sampling Methods to Estimate Probability of CYFIP1p Helical Conformation

Post date: May 20, 2018 6:58:45 PM by Megan Wang

Previously, we made modifications to the wild-type CYFIP1p to stabilize its α-helical structure and increase its binding efficiency to EIF4E. In this study, we utilized umbrella sampling methods to accelerate the sampling of possible peptide conformations, such that the probability of these conformations could be readily calculated from an MD simulation that would otherwise have taken months to collect a sufficient number of relevant samples. Using two Python scripts, we were able to apply a bias force onto CYFIP1p and measure the proportion of instances that CYFIP1p maintained a helical conformation. We effectively estimated the probability of such an outcome to be 0.18 percent if the bias force were nonexistent.

Potential bias force applied to interacting atoms in molecular system.  

Research Report

Scripts:

analysis.py

bias_MD.py