How to place a metal ion in protein model
Hi Its my pleasure to introduce myself, I am Suneel, graduate student from India. I am not a bioinformatician by profession but a eager aspirant of learning bioinformatics. I am new to Chimera, but as I started using it, the desire for leraning Bioinformatics increased more and more. Chimera is such an awesome tool packed with so many options. I love it . I have small question for you. I have a predicted the structure of protein using in silico tools (based on homology modeling). Literature sates that this protein binds to Magnesium. Now I would like to know if I can use chimera to place the metal in the modeled structure and prove/validate it based on metal geometry. Please do the needful. Thanking you in anticipation yours sincerely suneel
Hi Suneel, Chimera doesn't have a tool to predict metal binding sites from a structure. There is a tool called "Add Ions" which will roughly place ions around a structure, but it is meant for adding counterions for simulation purposes and is ineffective for binding site prediction.
First, I would consider the following:
(a) what protein was used as the template for the homology model? Does it bind metal? If so, the site is probably in the same place. You could just superimpose the template (with metal ion included) and model structures to see the binding site. http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/superposition.html
(b) do you have any other data about which residues bind the metal? You may be able to create a metal ion with Build Structure (under Tools... Structure Editing) and place the metal near the protein yourself, interactively, with the mouse. http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/editing/editing.html http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/mouse.html#activedef
(c) I believe there are several web servers that will try to identify metal-binding sites and other active sites from your uploaded protein structure. I would recommend doing some literature or web searching to find these. I have used some in the past, but it was a few years ago. Here are some web server names to search for and try: PAR-3D, PDBSiteScan, PINTS, ProFunc, Protemot, SuMo
Chimera has a Metal Geometry tool (under Tools... Structure Analysis), but it is meant to analyze the binding geometry of structures that already contain the metal ion, or to analyze the possibilities after you have manually placed the metal ion. http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/metalgeom/metalgeom.html
Best, Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 2, 2012, at 1:36 AM, Suneel N.A. wrote:
Hi Its my pleasure to introduce myself, I am Suneel, graduate student from India. I am not a bioinformatician by profession but a eager aspirant of learning bioinformatics. I am new to Chimera, but as I started using it, the desire for leraning Bioinformatics increased more and more. Chimera is such an awesome tool packed with so many options. I love it . I have small question for you. I have a predicted the structure of protein using in silico tools (based on homology modeling). Literature sates that this protein binds to Magnesium. Now I would like to know if I can use chimera to place the metal in the modeled structure and prove/validate it based on metal geometry. Please do the needful. Thanking you in anticipation yours sincerely suneel
Hi Elaine Thank you very much for the timely reply. I am glad to see it. The information you sent was really helpful. Here are some of the details for your suggestions/queries (a) I used 1ZWX for homology modelling. The other closest models are 2DDR and 3I5V. They are crystallised with bound metal ions.As you suggested I will try to superimpose with the metal ion (though I am not that good in using chimera).
I have another query: I am trying to model a protein based on homology modelling strategy. I came to know that Chimera can model the protein based on the same strategy. When I attempted modelling using the Modeller tool of chimera it gives me an error "Running modeller for query via web service failed; see reply log for more information. When I see the reply log : Import modeller, Import error:No module named modeller". Infact I even got the licence key from the modeller website and used it in the dialogue box but still this error message is shown.
Another query is, for modelling a protein (whose structure is unknown) , how do we import the sequences into chimera? In one of the chimera videos on modelling, they show that the protein sequence is imported/fetched from Uniport. But in my chimera software under the Fetch tab I don't see an Uniprot option. In this regard please tell me how to import sequences.
Please do the needful. Will be looking for the reply. Thanking you in anticipation Yours Sincerely suneel
On Wed, Jan 4, 2012 at 10:21 AM, Elaine Meng meng@cgl.ucsf.edu wrote:
Hi Suneel, Chimera doesn't have a tool to predict metal binding sites from a structure. There is a tool called "Add Ions" which will roughly place ions around a structure, but it is meant for adding counterions for simulation purposes and is ineffective for binding site prediction.
First, I would consider the following:
(a) what protein was used as the template for the homology model? Does it bind metal? If so, the site is probably in the same place. You could just superimpose the template (with metal ion included) and model structures to see the binding site. http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/superposition.html
(b) do you have any other data about which residues bind the metal? You may be able to create a metal ion with Build Structure (under Tools... Structure Editing) and place the metal near the protein yourself, interactively, with the mouse. < http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/editing/editing.htm...
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/mouse.html#activedef
(c) I believe there are several web servers that will try to identify metal-binding sites and other active sites from your uploaded protein structure. I would recommend doing some literature or web searching to find these. I have used some in the past, but it was a few years ago. Here are some web server names to search for and try: PAR-3D, PDBSiteScan, PINTS, ProFunc, Protemot, SuMo
Chimera has a Metal Geometry tool (under Tools... Structure Analysis), but it is meant to analyze the binding geometry of structures that already contain the metal ion, or to analyze the possibilities after you have manually placed the metal ion. < http://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/metalgeom/metalgeom...
Best, Elaine
Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 2, 2012, at 1:36 AM, Suneel N.A. wrote:
Hi Its my pleasure to introduce myself, I am Suneel, graduate student from
India. I am not a bioinformatician by profession but a eager aspirant of learning bioinformatics. I am new to Chimera, but as I started using it, the desire for leraning Bioinformatics increased more and more. Chimera is such an awesome tool packed with so many options. I love it . I have small question for you.
I have a predicted the structure of protein using in silico tools (based
on homology modeling). Literature sates that this protein binds to Magnesium. Now I would like to know if I can use chimera to place the metal in the modeled structure and prove/validate it based on metal geometry.
Please do the needful. Thanking you in anticipation yours sincerely suneel
Hi Suneel, The UniProt fetch is fairly new, you would need to get a recent daily build: http://www.cgl.ucsf.edu/chimera/download.html#daily http://www.cgl.ucsf.edu/chimera/docs/relnotes/snapshot.html
In older (but not too old) versions another approach is to get a file of the sequence in fasta format, then open it with "File... Open" in the Chimera menu.
If you get a newer Chimera and still have the modelling error, please use "Help... Report a Bug" in the menu to report it -- it will include information on your computer and what version you are using. This chimera-users email address is mainly for questions, not bugreports. Thanks!
There are explanations of how to superimpose structures (such as your template and model to see where the metal site would be) here: http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/superposition.html
Good luck, Elaine ---------- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco
On Jan 5, 2012, at 2:00 PM, Suneel N.A. wrote:
Hi Elaine Thank you very much for the timely reply. I am glad to see it. The information you sent was really helpful. Here are some of the details for your suggestions/queries (a) I used 1ZWX for homology modelling. The other closest models are 2DDR and 3I5V. They are crystallised with bound metal ions.As you suggested I will try to superimpose with the metal ion (though I am not that good in using chimera).
I have another query: I am trying to model a protein based on homology modelling strategy. I came to know that Chimera can model the protein based on the same strategy. When I attempted modelling using the Modeller tool of chimera it gives me an error "Running modeller for query via web service failed; see reply log for more information. When I see the reply log : Import modeller, Import error:No module named modeller". Infact I even got the licence key from the modeller website and used it in the dialogue box but still this error message is shown.
Another query is, for modelling a protein (whose structure is unknown) , how do we import the sequences into chimera? In one of the chimera videos on modelling, they show that the protein sequence is imported/fetched from Uniport. But in my chimera software under the Fetch tab I don't see an Uniprot option. In this regard please tell me how to import sequences.
Please do the needful. Will be looking for the reply. Thanking you in anticipation Yours Sincerely suneel
Hi:
I have given my students a little problem that involves producing an "animation" with each step in the animation being a PDB model within an NMR generated file. The solution to the problem is more complicated than what is shown in the script below involving as it does the calculation of RMSDs to get smoother transitions between successive displays of the models.
My answer script is not working as expected due to an issue that I have recreated in a shorter script that follows next:
import chimera from chimera import runCommand import time
numCycles = 10
# ID's to try: 1BMR, 1DW5, 1QDP, 1D1H, 1N4N, 1IJC, 1PFD, 2LI3, 1AH1 prot_L = chimera.openModels.open("1AH1", type="PDB") numModels = len(prot_L)
runCommand("~modeldisp")
for ic in range(numCycles): for aix in range(numModels): subIDstr = str(prot_L[aix].subid) command = "modeldisp #0" + "." + subIDstr runCommand(command) time.sleep(0.1) command = "~modeldisp #0" + "." + subIDstr # The following print command seems to be necessary to get # the animation! print command runCommand(command)
This script runs just fine and does exactly what I wanted it to do. Here is the problem: If I remove the print statement, initially there for debugging, the script does not work! It seems weird that printing the argument for the runCommand would have any effect on the execution of the runCommand statement. Am I missing some vital point here????
I am running the latest version of Chimera: (version 1.6 build 35398) under Windows 7 Professional. Prior to each run I have being doing
File... Close Session
to eliminate any models that were brought in on a previous run.
Cheers, Forbes Burkowski
Hi Forbes, The trick is that, unlike a Chimera command file, there is no implicit "wait 1" (i.e. show a new frame) at the end of commands executed with the Python runCommand call. Therefore with this sequence of runCommands:
runCommand("color red") runCommand("color green") runCommand("color blue")
the display will never show the models colored red or green, just blue. So, with the print command removed from your script the modeldisp/~modeldisp commands are executed "back to back" and you never see anything. Though due to the arcane minutiae of how the print command works, you do wind up getting a frame redraw after a print which is why it works with the print in there. Also, Chimera does not execute at all (including frame redraws) during your time.sleep(0.1). To get the script to work the way you want your options are:
1) add "; wait 1" to the end of the modeldisp commands 2) change time.sleep(0.1) to runCommand("wait 3") [3 frames at 30 frames/second] 3) both
Sorry for the inscrutableness of all this.
--Eric
On Feb 6, 2012, at 7:10 PM, Forbes J. Burkowski wrote:
Hi:
I have given my students a little problem that involves producing an "animation" with each step in the animation being a PDB model within an NMR generated file. The solution to the problem is more complicated than what is shown in the script below involving as it does the calculation of RMSDs to get smoother transitions between successive displays of the models.
My answer script is not working as expected due to an issue that I have recreated in a shorter script that follows next:
import chimera from chimera import runCommand import time
numCycles = 10
# ID's to try: 1BMR, 1DW5, 1QDP, 1D1H, 1N4N, 1IJC, 1PFD, 2LI3, 1AH1 prot_L = chimera.openModels.open("1AH1", type="PDB") numModels = len(prot_L)
runCommand("~modeldisp")
for ic in range(numCycles): for aix in range(numModels): subIDstr = str(prot_L[aix].subid) command = "modeldisp #0" + "." + subIDstr runCommand(command) time.sleep(0.1) command = "~modeldisp #0" + "." + subIDstr # The following print command seems to be necessary to get # the animation! print command runCommand(command)
This script runs just fine and does exactly what I wanted it to do. Here is the problem: If I remove the print statement, initially there for debugging, the script does not work! It seems weird that printing the argument for the runCommand would have any effect on the execution of the runCommand statement. Am I missing some vital point here????
I am running the latest version of Chimera: (version 1.6 build 35398) under Windows 7 Professional. Prior to each run I have being doing
File... Close Session
to eliminate any models that were brought in on a previous run.
Cheers, Forbes Burkowski
Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
Hi:
Thanks for the advice. I'll have to determine some reasonable way to communicate this to my class. I do have another problem though!
I have a rather long script that involves some VRML and it is also misbehaving. The script to follow has the same behaviour and is much simpler:
import chimera from chimera import runCommand import CGLutil from CGLutil import vrml def vr_plot(): wrl = vrml.Transform() for ii in range(10): for jj in range(10): translateBaseCube = vrml.Transform(translation = tuple([jj, ii, 0])) base = vrml.Box(size = (.1, .1, .1), color = (1.,1.,.5)) translateBaseCube.addChild(base) wrl.addChild(translateBaseCube) return vrml.vrml(wrl) vvrr = vr_plot() vrlmModel = chimera.openModels.open(vvrr, type = 'VRML')[0] chimera.viewer.viewAll()
This VRML snippet gives me an array of 100 little cubes. Here's the catch:
On last year's version of Chimera (1.5.3) the script generates the scene almost instantaneously. With the current version of Chimera (the January release or the Feb. 04 release) the execution is extremely slow. I am running Windows 7 on a machine with 12 GB of RAM and an Intel i7 CPU 3.2GHz so I am not lacking for computing power. One of my students also tried this script and reports a similar slow down. I am also noticing that the latest version seems to launch more slowly and takes more time to display BIG scenes (consider PDB ID = 1FFK).
Is there some way to improve this situation or do I have to give up on using VRML?
Cheers, Forbes
On Tue, 7 Feb 2012, Eric Pettersen wrote:
Hi Forbes, The trick is that, unlike a Chimera command file, there is no implicit "wait 1" (i.e. show a new frame) at the end of commands executed with the Python runCommand call. Therefore with this sequence of runCommands:
runCommand("color red") runCommand("color green") runCommand("color blue")
the display will never show the models colored red or green, just blue. So, with the print command removed from your script the modeldisp/~modeldisp commands are executed "back to back" and you never see anything. Though due to the arcane minutiae of how the print command works, you do wind up getting a frame redraw after a print which is why it works with the print in there. Also, Chimera does not execute at all (including frame redraws) during your time.sleep(0.1). To get the script to work the way you want your options are:
- add "; wait 1" to the end of the modeldisp commands
- change time.sleep(0.1) to runCommand("wait 3") [3 frames at 30
frames/second] 3) both
Sorry for the inscrutableness of all this.
--Eric
On Feb 6, 2012, at 7:10 PM, Forbes J. Burkowski wrote:
Hi:
I have given my students a little problem that involves producing an "animation" with each step in the animation being a PDB model within an NMR generated file. The solution to the problem is more complicated than what is shown in the script below involving as it does the calculation of RMSDs to get smoother transitions between successive displays of the models.
My answer script is not working as expected due to an issue that I have recreated in a shorter script that follows next:
import chimera from chimera import runCommand import time
numCycles = 10
# ID's to try: 1BMR, 1DW5, 1QDP, 1D1H, 1N4N, 1IJC, 1PFD, 2LI3, 1AH1 prot_L = chimera.openModels.open("1AH1", type="PDB") numModels = len(prot_L)
runCommand("~modeldisp")
for ic in range(numCycles): for aix in range(numModels): subIDstr = str(prot_L[aix].subid) command = "modeldisp #0" + "." + subIDstr runCommand(command) time.sleep(0.1) command = "~modeldisp #0" + "." + subIDstr # The following print command seems to be necessary to get # the animation! print command runCommand(command)
This script runs just fine and does exactly what I wanted it to do. Here is the problem: If I remove the print statement, initially there for debugging, the script does not work! It seems weird that printing the argument for the runCommand would have any effect on the execution of the runCommand statement. Am I missing some vital point here????
I am running the latest version of Chimera: (version 1.6 build 35398) under Windows 7 Professional. Prior to each run I have being doing
File... Close Session
to eliminate any models that were brought in on a previous run.
Cheers, Forbes Burkowski
Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
On Feb 7, 2012, at 1:07 PM, Forbes J. Burkowski wrote:
This VRML snippet gives me an array of 100 little cubes. Here's the catch:
On last year's version of Chimera (1.5.3) the script generates the scene almost instantaneously. With the current version of Chimera (the January release or the Feb. 04 release) the execution is extremely slow. I am running Windows 7 on a machine with 12 GB of RAM and an Intel i7 CPU 3.2GHz so I am not lacking for computing power. One of my students also tried this script and reports a similar slow down.
Urg. You're right; whereas it takes a couple of seconds to open on 1.5.3, it takes basically forever on 1.6. I'll be opening a ticket in our bug database about this with you on the recipient list so you'll know when it is fixed. I hope it gets fixed in time for the 1.6 final version but if not then it will be fixed in 1.6.1 for sure.
I am also noticing that the latest version seems to launch more slowly and takes more time to display BIG scenes (consider PDB ID = 1FFK).
Well, this is actually a side effect of the VRML thing. I hadn't anticipated the direct opening of VRML strings when writing the new Rapid Access interface. Consequently, the text of one of the "recently opened" buttons is the entirety of the VRML string you opened! It actually only shows the last two curly braces, LOL. At any rate, showing that huge VRML string on the button is causing all the slowdown.
I have committed a fix for this latter problem and it will be in the next successful build. If the slowness is bugging you too much to wait you could go and edit ~/.chimera/preferences, search for "Rapid Access", find the first opening curly brace after it (start of a dictionary) and change the 'data history' list to an empty list.
Thanks for reporting this, BTW.
--Eric
Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu
On Feb 7, 2012, at 2:44 PM, Eric Pettersen wrote:
I have committed a fix for this latter problem and it will be in the next successful build. If the slowness is bugging you too much to wait you could go and edit ~/.chimera/preferences, search for "Rapid Access", find the first opening curly brace after it (start of a dictionary) and change the 'data history' list to an empty list.
It turns out that this fix also fixes the slow-VRML-open thing. I've attached a patch that you could apply to <your Chimera installation>/ share/chimera/__init__.py if you don't want to wait for the next build.
--Eric
Hi: Thanks for looking into this.
By the way, I just happened to chance upon the Protein Data Bank page in Wikipedia:
http://en.wikipedia.org/wiki/Protein_Data_Bank
If you go down to the section on "Viewing the data" there is a lengthy list of molecular display programs including Chimera. They all have links to the homepages except for UCSF Chimera!
This nasty omission should be corrected as soon as possible.
Cheers, Forbes
On Tue, 7 Feb 2012, Eric Pettersen wrote:
On Feb 7, 2012, at 1:07 PM, Forbes J. Burkowski wrote:
This VRML snippet gives me an array of 100 little cubes. Here's the catch:
On last year's version of Chimera (1.5.3) the script generates the scene almost instantaneously. With the current version of Chimera (the January release or the Feb. 04 release) the execution is extremely slow. I am running Windows 7 on a machine with 12 GB of RAM and an Intel i7 CPU 3.2GHz so I am not lacking for computing power. One of my students also tried this script and reports a similar slow down.
Urg. You're right; whereas it takes a couple of seconds to open on 1.5.3, it takes basically forever on 1.6. I'll be opening a ticket in our bug database about this with you on the recipient list so you'll know when it is fixed. I hope it gets fixed in time for the 1.6 final version but if not then it will be fixed in 1.6.1 for sure.
I am also noticing that the latest version seems to launch more slowly and takes more time to display BIG scenes (consider PDB ID = 1FFK).
Well, this is actually a side effect of the VRML thing. I hadn't anticipated the direct opening of VRML strings when writing the new Rapid Access interface. Consequently, the text of one of the "recently opened" buttons is the entirety of the VRML string you opened! It actually only shows the last two curly braces, LOL. At any rate, showing that huge VRML string on the button is causing all the slowdown.
I have committed a fix for this latter problem and it will be in the next successful build. If the slowness is bugging you too much to wait you could go and edit ~/.chimera/preferences, search for "Rapid Access", find the first opening curly brace after it (start of a dictionary) and change the 'data history' list to an empty list.
Thanks for reporting this, BTW.
--Eric
Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu
Hi Forbes, Thanks for letting us know. I made it link to the "UCSF Chimera" wikipedia entry. Sigh, nowadays it is not enough just to keep your own website updated, all these other places need monitoring too! Elaine ----- Elaine C. Meng, Ph.D. UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco
On Feb 7, 2012, at 4:23 PM, Forbes J. Burkowski wrote:
Hi: Thanks for looking into this.
By the way, I just happened to chance upon the Protein Data Bank page in Wikipedia:
http://en.wikipedia.org/wiki/Protein_Data_Bank
If you go down to the section on "Viewing the data" there is a lengthy list of molecular display programs including Chimera. They all have links to the homepages except for UCSF Chimera!
This nasty omission should be corrected as soon as possible.
Cheers, Forbes
Forgive me for asking an obvious (to you) question,,, when I label/other to place, for example, spin quantities on each atom- where can I change the font and color of these labels so that they are more easily read ?? Thanks
Barry
Hi Barry,
To change the label color use menu
Actions / Color / All Options...
and choose the Color Applies To "atom labels" option to color just the labels of the selected atoms. Then click on the color you want. To adjust the font size use menu entry
Favorites / Preferences
choose Category "Background", and you can adjust the Label font size there. All atom labels have to have the same size.
There are also "2d" labels that aren't attached to the atoms under
Tools / Utilities / 2D Labels
and those can be independently sized and colored. But they don't move when you rotate your molecule. Can be useful for making presentation images though.
Tom
Forgive me for asking an obvious (to you) question,,, when I label/other to place, for example, spin quantities on each atom- where can I change the font and color of these labels so that they are more easily read ?? Thanks
Barry
participants (6)
-
Blever -
Elaine Meng -
Eric Pettersen -
Forbes J. Burkowski -
Suneel N.A. -
Tom Goddard