Dear Amber-Users, Dear Amber-Developers
I am looking at the Generalized Born-source-code and to me (and one of
my supervisors) it seems there is a coding-mistake in the file
$AMBERHOME/AmberTools/src/sander/egb.F90 (attached)
I am referring to "AmberTools version 18.08" (amber18; "sander: Version
17.0"):
In Line 750 "vectmp3(1:icount) = r2x(1:icount) +
rjx(1:icount)*ri*vectmp1(1:icount)" which is fij, which seems to be
correct, according to Equation (4.3) of the Amber-Manual 18.
"vectmp1(1:icount)" is defined in Line 743 as "call vdexp( icount,
vectmp1, vectmp1 )", with "vectmp1(1:icount) =
-r2x(1:icount)*vectmp1(1:icount)", defined in line 742, which seems to
be wrong.
Accoding to Manual-Equation (4.3) the term in exp() is dimensionless:
-r_{ij}^2/(4R_iR_j), which seems to make sence.
Hence it should be "vectmp1(1:icount) =
-r2x(1:icount)**1./*vectmp1(1:icount)", since vectmp1 is defined in Line
738&710&700&515 as "vectmp1(j-i) = four*reff(i)*reff(j)".
Best Regards
Johannes
--------------
Source-Code-lines for "AmberTools version 17.10" (amber16; "sander:
Version 17.0"):
In Line 745 "vectmp3(1:icount) = r2x(1:icount) +
rjx(1:icount)*ri*vectmp1(1:icount)" which is fij, which seems to be
correct, according to Equation (4.3) of the Amber-Manual 16.
"vectmp1(1:icount)" is defined in Line 738 as "call vdexp( icount,
vectmp1, vectmp1 )", with "vectmp1(1:icount) =
-r2x(1:icount)*vectmp1(1:icount)",defined in line 737, which seems to be
wrong.
Accoding to Manual-Equation (4.3) the term in exp() is dimensionless:
-r_{ij}^2/(4R_iR_j), which seems to make sence.
Hence it should be "vectmp1(1:icount) =
-r2x(1:icount)**1./*vectmp1(1:icount)", since vectmp1 is defined in Line
733&705&695&510 as "vectmp1(j-i) = four*reff(i)*reff(j)".
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Aug 09 2018 - 04:30:01 PDT
