Querries on eABF

From: Souvik Sinha (souvik.sinha893_at_gmail.com)
Date: Wed Apr 26 2017 - 01:40:21 CDT

I was trying eABF (on 1D) with the example system given within NAMD 2.12
package. It seems to be generating pmf files directly along with '.grad' &
'.count' without any post-processing script. Maybe that's the whole point
of "on-the-fly" implementation of eABF. But there are 2 files for
pmf/grad/count. Along with normal '(abf).grad/.pmf', there are
'(abf.czar).pmf/.grad'. And along with normal '.count', there is a
'.zcout'. I believe these normal .grad/.pmf is the output of Zhang-Yang
estimator & .czar.* is of CZAR estimator. Am I right? And what's this
.zcount ?

Then I tried 2D with the NANMA system (provided with ABF tutorial files).
Now, whenever I am trying to merge '.grad' & '.count' of individual
windows using 'colvar-merge.in' , its giving the following error:

" colvars: Error opening eABF z-gradient file output/colvarBo.zgrad for
reading "

There aren't any '.zgrad' file generated as an output of any eABF run.
Again only '.czar.grad' files are there. If these '.zgrad' & '.czar.grad'
are same then I can rename & proceed further. But I am not certain. Please
enlighten me.

And finally, is multiple-walker algorithm applicable for 2D using eABF?

Looking for useful suggestions. Thanks.

Souvik Sinha
Research Fellow
Bioinformatics Centre (SGD LAB)
Bose Institute
Contact: 033 25693275

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:14 CST