Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models. When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction. Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved. thanks, Matt
Hi Matt, Yeah, the savepos/reset commands is pretty unfriendly. I think the sensible fix is that "reset" should be changed to move unknown models in some sensible way instead of just leaving them where they lie. The common case is that the transformation of all remembered models is the same, so position any new model using that same transformation. If some models have different saved transformations then it might be reasonable to look just at the models with the same current transformation as the new model -- if all those end up with the same transform then apply that to the new model as well. If there is some real ambiguity (models with the same current transform match the new model and then move to different relative positions in the saved position) then obviously you have to specify how to incorporate the new model in that saved position. I won't have time to work on this for probably a few months. Perhaps another Chimera developer will look at it. As things currently work I think the only approach is quite painful. You have to reset to a saved position then use the Model Panel / Match button to align your new model with one of the existing models, then reuse the "savepos" command with the original name to augment it. Ugh. The matrixcopy command does the same think as the Match button and may be easier. Tom Matthew Dougherty wrote:
Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models.
When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction.
Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved.
thanks, Matt
Hi Matt, I agree with most of what Tom said, except that "match" in the Model Panel is not the same as using "matrixcopy" - the "matrixcopy" command applies the current transformation of one model to another, whereas "match" in the Model Panel calls MatchMaker. MatchMaker constructs a sequence alignment and uses the pairings in the alignment to superimpose the structures. In the morphing case, it generally does not work to use "matrixcopy" after resetting the original models to some saved position, because the untransformed position of the morph corresponds to the transformed positions of the input models at the time the morph was created. For example, if you use "reset" without arguments to untransform all models right after creating the morph, the morph will stay put whereas the other models will go back to their starting positions, no longer sensible relative to the morph position. Initially I thought "matrixcopy" would work... then I actually tried it. 8-( That still leaves the other approach Tom described: reset to saved position, use MatchMaker (or matchmaker command) to superimpose the new model onto one of the old ones included in the saved position, re- save the position. I can imagine one variant, which may not be any better: after using matchmaker, write out the new model "relative to" the old model, close it, read it back in. Then after resetting to the various saved positions, you could use "matrixcopy" to apply the old model's transform to the new model, since their untransformed coordinates are now in sync. Perhaps all this babbling is incomprehensible - if so, I apologize! Elaine ----- Elaine C. Meng, Ph.D. meng@cgl.ucsf.edu UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco http://www.cgl.ucsf.edu/home/meng/index.html On Dec 8, 2008, at 11:08 PM, Tom Goddard wrote:
Hi Matt,
Yeah, the savepos/reset commands is pretty unfriendly. I think the sensible fix is that "reset" should be changed to move unknown models in some sensible way instead of just leaving them where they lie. The common case is that the transformation of all remembered models is the same, so position any new model using that same transformation. If some models have different saved transformations then it might be reasonable to look just at the models with the same current transformation as the new model -- if all those end up with the same transform then apply that to the new model as well. If there is some real ambiguity (models with the same current transform match the new model and then move to different relative positions in the saved position) then obviously you have to specify how to incorporate the new model in that saved position. I won't have time to work on this for probably a few months. Perhaps another Chimera developer will look at it.
As things currently work I think the only approach is quite painful. You have to reset to a saved position then use the Model Panel / Match button to align your new model with one of the existing models, then reuse the "savepos" command with the original name to augment it. Ugh. The matrixcopy command does the same think as the Match button and may be easier.
Tom
Matthew Dougherty wrote:
Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models.
When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction.
Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved.
thanks, Matt
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
In my previous email where I referred to the "Match" button of Model Panel I meant to say the "Transform As" button. The behavior Elaine describes for morphs sounds like a bug. A newly created morph model (trajectory) should have the same transformation as the first input model used to create the morph. From Elaine's description it sounds like the transform of the morph model is unrelated to the transforms of the input models making alignment when reopening the data sets problematic. Tom Elaine Meng wrote:
Hi Matt, I agree with most of what Tom said, except that "match" in the Model Panel is not the same as using "matrixcopy" - the "matrixcopy" command applies the current transformation of one model to another, whereas "match" in the Model Panel calls MatchMaker. MatchMaker constructs a sequence alignment and uses the pairings in the alignment to superimpose the structures.
In the morphing case, it generally does not work to use "matrixcopy" after resetting the original models to some saved position, because the untransformed position of the morph corresponds to the transformed positions of the input models at the time the morph was created. For example, if you use "reset" without arguments to untransform all models right after creating the morph, the morph will stay put whereas the other models will go back to their starting positions, no longer sensible relative to the morph position. Initially I thought "matrixcopy" would work... then I actually tried it. 8-(
That still leaves the other approach Tom described: reset to saved position, use MatchMaker (or matchmaker command) to superimpose the new model onto one of the old ones included in the saved position, re- save the position. I can imagine one variant, which may not be any better: after using matchmaker, write out the new model "relative to" the old model, close it, read it back in. Then after resetting to the various saved positions, you could use "matrixcopy" to apply the old model's transform to the new model, since their untransformed coordinates are now in sync.
Perhaps all this babbling is incomprehensible - if so, I apologize! Elaine ----- Elaine C. Meng, Ph.D. meng@cgl.ucsf.edu UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco http://www.cgl.ucsf.edu/home/meng/index.html
On Dec 8, 2008, at 11:08 PM, Tom Goddard wrote:
Hi Matt,
Yeah, the savepos/reset commands is pretty unfriendly. I think the sensible fix is that "reset" should be changed to move unknown models in some sensible way instead of just leaving them where they lie. The common case is that the transformation of all remembered models is the same, so position any new model using that same transformation. If some models have different saved transformations then it might be reasonable to look just at the models with the same current transformation as the new model -- if all those end up with the same transform then apply that to the new model as well. If there is some real ambiguity (models with the same current transform match the new model and then move to different relative positions in the saved position) then obviously you have to specify how to incorporate the new model in that saved position. I won't have time to work on this for probably a few months. Perhaps another Chimera developer will look at it.
As things currently work I think the only approach is quite painful. You have to reset to a saved position then use the Model Panel / Match button to align your new model with one of the existing models, then reuse the "savepos" command with the original name to augment it. Ugh. The matrixcopy command does the same think as the Match button and may be easier.
Tom
Matthew Dougherty wrote:
Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models.
When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction.
Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved.
thanks, Matt
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
Morph only uses the transformed coordinates for all included models because they may be manually aligned. It makes no distinction between the first and the others. Conrad Tom Goddard wrote:
In my previous email where I referred to the "Match" button of Model Panel I meant to say the "Transform As" button.
The behavior Elaine describes for morphs sounds like a bug. A newly created morph model (trajectory) should have the same transformation as the first input model used to create the morph. From Elaine's description it sounds like the transform of the morph model is unrelated to the transforms of the input models making alignment when reopening the data sets problematic.
Tom
Elaine Meng wrote:
Hi Matt, I agree with most of what Tom said, except that "match" in the Model Panel is not the same as using "matrixcopy" - the "matrixcopy" command applies the current transformation of one model to another, whereas "match" in the Model Panel calls MatchMaker. MatchMaker constructs a sequence alignment and uses the pairings in the alignment to superimpose the structures.
In the morphing case, it generally does not work to use "matrixcopy" after resetting the original models to some saved position, because the untransformed position of the morph corresponds to the transformed positions of the input models at the time the morph was created. For example, if you use "reset" without arguments to untransform all models right after creating the morph, the morph will stay put whereas the other models will go back to their starting positions, no longer sensible relative to the morph position. Initially I thought "matrixcopy" would work... then I actually tried it. 8-(
That still leaves the other approach Tom described: reset to saved position, use MatchMaker (or matchmaker command) to superimpose the new model onto one of the old ones included in the saved position, re- save the position. I can imagine one variant, which may not be any better: after using matchmaker, write out the new model "relative to" the old model, close it, read it back in. Then after resetting to the various saved positions, you could use "matrixcopy" to apply the old model's transform to the new model, since their untransformed coordinates are now in sync.
Perhaps all this babbling is incomprehensible - if so, I apologize! Elaine ----- Elaine C. Meng, Ph.D. meng@cgl.ucsf.edu UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco http://www.cgl.ucsf.edu/home/meng/index.html
On Dec 8, 2008, at 11:08 PM, Tom Goddard wrote:
Hi Matt,
Yeah, the savepos/reset commands is pretty unfriendly. I think the sensible fix is that "reset" should be changed to move unknown models in some sensible way instead of just leaving them where they lie. The common case is that the transformation of all remembered models is the same, so position any new model using that same transformation. If some models have different saved transformations then it might be reasonable to look just at the models with the same current transformation as the new model -- if all those end up with the same transform then apply that to the new model as well. If there is some real ambiguity (models with the same current transform match the new model and then move to different relative positions in the saved position) then obviously you have to specify how to incorporate the new model in that saved position. I won't have time to work on this for probably a few months. Perhaps another Chimera developer will look at it.
As things currently work I think the only approach is quite painful. You have to reset to a saved position then use the Model Panel / Match button to align your new model with one of the existing models, then reuse the "savepos" command with the original name to augment it. Ugh. The matrixcopy command does the same think as the Match button and may be easier.
Tom
Matthew Dougherty wrote:
Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models.
When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction.
Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved.
thanks, Matt
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
Hi Matt, I've updated 'reset' so that new models keep their relative position to the lowest-numbered old model when resetting to a position that didn't include the new model. I hope that helps some. It'll be in tonight's daily build. In the long term I intend to implement an "updatepos" command where you can specify which model a new model should maintain its position relative to. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu On Dec 8, 2008, at 5:58 PM, Matthew Dougherty wrote:
Let's say I have 5 models and I save a number of positions (savepos). Then I later create some morphs between the models.
When I do a reset of a position, the morph model's position does not correspond with the other model positions. This is partially due to the method of camera control in chimera of moving the models and keeping the camera position pointing in the z direction.
Is there a way to incorporate the morph model into the saved positions that were created prior to the creation of the morph model? having similar problem with models I open after the positions are saved.
thanks, Matt _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users
participants (5)
-
Conrad Huang -
Elaine Meng -
Eric Pettersen -
Matthew Dougherty -
Tom Goddard