Hi all,

I am encountering a strange inconsistency with the 'relative' option for the "write" command. Searching for similar problems on the mailing list came up short, so I thought I'd ask about it.

To put it briefly, I run chimera with --nogui and perform ~780,000 total matches in which I write out new coordinates using 'relative.' My issue lies in automating this process with .com scripts, which creates different results from typing out the exact same commands into the command line of the GUI. This has been plaguing me for a year or so, and most recently I can cite using version 1.10.1 (build 40427).

Below is a small example to illustrate my operations. The "B-new_#.pdb" output files will contain different coordinates if they originate from automated runs or manual execution.

This problem is sporadic and I cannot narrow down the source of the discrepancy. From best I can tell, there are occasionally output files from automation (e.g. B-new_#.pdb) that are in the correct coordinate reference frame. However, many (most?) are not and it is a big obstacle.

I hope that someone may have some insight on this problem.

/////////// example ///////////
open A.pdb
open B.pdb
open C.pdb

##write out selection from #0 [this file is always consistent]
select #0:...
write selected relative 0 #0 A-new.pdb

##move #2 onto #0
match #2:... #0:...
##move #1 onto #2
match #1:... #2:...

##write out selection from #1 and save it in its new position relative to #0 [inconsistent]
select #1:...
write selected relative 0 #1 B-new_1.pdb
~select all

##move #1 slightly
match #1:... #2:...+1

##write out selection from #1 and save it in its new position relative to #0 [inconsistent]
select #1:...
write selected relative 0 #1 B-new_2.pdb
~select all

##move #1 slightly more
match #1:... #2:...+2

##write out selection from #1 and save it in its new position relative to #0 [inconsistent]
select #1:...
write selected relative 0 #1 B-new_3.pdb
/////////// example ///////////

Best,
Dan McNamara