[Dock-fans] 6.2 bugfix 1
Scott Brozell
sbrozell at scripps.edu
Tue May 6 00:01:09 PDT 2008
http://dock.compbio.ucsf.edu/DOCK_6/bugfixes.htm
********>Bugfix 1:
Author: Scott Brozell
Date: May 5, 2008
Program: dock
Description: the Amber score distance movable residue selection was
not correct. This bug was only exercised when a strand had no residues
within the distance cutoff; in that case the whole strand instead of none
of the strand was selected by the generated NAB atom expression.
Fix: make the following changes to the file src/dock/score_amber.cpp
-------------------------------------------------------------------------
*** src/dock/score_amber.cpp 19 Dec 2007 18:58:06 -0000 1.75
--- src/dock/score_amber.cpp 6 May 2008 00:01:47 -0000 1.77
***************
*** 743,750 ****
// NAB idiom to loop over all atoms
for( n = 0, sp = (*m)->m_strands; sp; sp = sp->s_next ){
! residue_numbers.insert( ResiduesPerStrand::value_type(
! sp->s_strandname, set< Residue_Sequence_Number >() ) );
for( r = 0; r < sp->s_nresidues; r++ ){
res = sp->s_residues[ r ];
for( a = 0; a < res->r_natoms; a++ ){
--- 743,749 ----
// NAB idiom to loop over all atoms
for( n = 0, sp = (*m)->m_strands; sp; sp = sp->s_next ){
! set< Residue_Sequence_Number> rsnset;
for( r = 0; r < sp->s_nresidues; r++ ){
res = sp->s_residues[ r ];
for( a = 0; a < res->r_natoms; a++ ){
***************
*** 755,761 ****
ap->a_pos[ 2 ] );
if ( i->crds.distance_squared( axyz ) <
distance_cutoff*distance_cutoff ) {
! residue_numbers[sp->s_strandname].insert( r + 1 );
++ n;
goto NEXT_RESIDUE; // exit sphere and atom loops
}
--- 754,760 ----
ap->a_pos[ 2 ] );
if ( i->crds.distance_squared( axyz ) <
distance_cutoff*distance_cutoff ) {
! rsnset.insert( r + 1 );
++ n;
goto NEXT_RESIDUE; // exit sphere and atom loops
}
***************
*** 764,769 ****
--- 763,772 ----
NEXT_RESIDUE:
; // null statement
}
+ if ( ! rsnset.empty() ) {
+ residue_numbers.insert( ResiduesPerStrand::value_type(
+ sp->s_strandname, rsnset ) );
+ }
}
ostringstream mae;
--------------------------------------------------------------------------
Workaround: none.
More information about the Dock-fans
mailing list