From: Jerome Henin (jhenin_at_cmm.chem.upenn.edu)
Date: Wed Aug 12 2009 - 17:49:24 CDT
I have just found a bug in the integration tool for multidimensional
ABF calculations (abf_integrate). It will affect simulations where
only a small area is sampled, surrounded by a region where no sampling
at all is collected. In that case, the integration tool uses incorrect
boundary conditions, leading to some amount of bias in the integrated
PMF. An easy symptom to notice is that the final PMF is not uniform in
the unsampled regions.
The problem has been corrected, and new source files will be made
available shortly on the public CVS server.
Sorry for the inconvenience,
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:51:22 CST