MantisBT - Rosetta
View Issue Details
0000097Rosetta[All Projects] Bad Codingpublic2012-09-06 12:412017-08-07 20:02
smlewis 
amw579 
lowminoralways
closedwon't fix 
 
 
not an application
not related to a command line
n/a
0000097: Lysine (LYS.params and derivatives) has ridiculous starting conformation; may be true for other residues as well
Instantiating a lysine residue from the params file results in a ridiculous eclipsed structure that looks more like a 5-membered ring, where the tip of the lysine overlays the backbone. Why not change the params file to have a more reasonable set of starting torsions (a common fully extended rotamer)?
Switch a centroid lysine to fullatom with SwitchToResidueTypeSet without repacking, or instantiate a lysine residue from scratch (like with make_pose_from_sequence) and dump to PDB
I bet this will cause test changes all over the place...
No tags attached.
Issue History
2012-09-06 12:41smlewisNew Issue
2012-09-09 21:48LabonteNote Added: 0000094
2012-09-09 21:49LabonteAssigned To => Labonte
2012-09-09 21:49LabonteStatusnew => confirmed
2012-09-09 21:51LabonteAssigned ToLabonte =>
2013-02-26 12:44LabonteNote Added: 0000144
2013-02-28 14:51smlewisNote Added: 0000147
2013-02-28 14:52smlewisNote Edited: 0000147bug_revision_view_page.php?bugnote_id=147#r51
2013-03-07 09:26smlewisNote Added: 0000161
2017-08-07 20:02amw579Fixed in SVN Version => n/a
2017-08-07 20:02amw579Statusconfirmed => closed
2017-08-07 20:02amw579Assigned To => amw579
2017-08-07 20:02amw579Resolutionopen => won't fix

Notes
(0000094)
Labonte   
2012-09-09 21:48   
Yeah, they are all like that. The omega values are set to 0, for example. I've been bothered by this since I first joined Rosetta. But when I brought it up at the last MiniCON, some people indicated that someone had a good reason for not changing it. I'd love to change it, if everyone agrees with the idea.
(0000144)
Labonte   
2013-02-26 12:44   
So, can I have permission on this one too?
(0000147)
smlewis   
2013-02-28 14:51   
(edited on: 2013-02-28 14:52)
It didn't get brought up at minicon...we should take it to the list and see if we can shake loose the reason we weren't supposed to change it, so we can note it here and close as unresolvable...or alternately fix it.

(0000161)
smlewis   
2013-03-07 09:26   
Discussion on the list came to no consensus.

Arguments in favor of the change:
-do not knowingly place nonsense rotamers that clash with their own backbone, --this confuses new/casual users
--it's ugly
--it's wrong, in the sense of we know those rotamers cannot ever be right

Arguments against:
-Obviously bad rotamers signal that the structure hasn't been properly repacked and needs to be
--Non-obviously bad rotamers will fail to signal so clearly that something was forgotten
--New users are upset by bad rotamers but respond with packing; not-so-bad rotamers might not be noticed and might go unrepacked

I think we are on track to close this as unresolvable.