VMD-L Mailing List

From: Kevin Song (kcsong_at_utexas.edu)
Date: Tue Oct 23 2018 - 11:58:47 CDT

Hello all,

I'm trying to use parseFEP to analyze the result of a protein mutation
experiment. I'm using files/instructions from the FEP tutorial, modified
slightly to fit my needs. The NAMD simulations seem to work correctly,
but when I go to analyze the results, I get the following error:

domain error: argument not in valid range
 
 

domain error: argument not in valid range
     while executing
"expr $var_now/$f < $t +1 && $var_now / $f > 1 - $t "
     (procedure "::ParseFEP::FEP_formula" line 196)
     invoked from within
"::ParseFEP::FEP_formula $file $file_entropy $window $mean_xi $fororback"
     (procedure "::ParseFEP::normal_parse_log" line 55)
     invoked from within
     <rest of stack trace>

Does anyone know what I might be doing wrong?

I've checked the following to rule out some common issues:
    - xmgrace and imagemagick tools are installed and on the PATH
    - FEP is being run in a symmetric manner between backward and forward
      simulations:
         runFEP 0.0 1.0 0.025 $numSteps
        runFEP 1.0 0.0 -0.025 $numSteps

I'm fairly new to MD calculations in general, so I might have just
missed out on some obvious setting or flag I needed to toggle. My NAMD
configs, fepout files, and log files are available at
https://utexas.box.com/shared/static/1yvad1ca4qkn41yr5qgojxp1dlwg5oyt.gz

I'd appreciate any pointers to how I could fix this!

P.S. I've added a few puts statements into the parseFEP script and
confirmed that the problem is that $f is zero, but I have no idea why
that might be the case. 

-- 
*KEVIN SONG*, Graduate Student
The University of Texas at Austin | Computer Science | 408-657-6646 |
cs.utexas.edu