PCA and traj
Issue #780
new
What’s good practice on choosing the trajectory data (full trajectory, 0-100ns, or just last 10ns)?
What’s good practice on choosing the trajectory data (full trajectory, 0-100ns, or just last 10ns)?
Hi,
The answer to this question is really system dependent. Generally, if a simulation reaches some kind of “equilibrium” (e.g., measured by backbone RMSD of the system), either taking the “equilibrated” part or the whole trajectory should be fine. Also, 100ns seems too short according to the current standard (nowadays, microsecond simulations can be routinely performed with some descent GPU).