[Chimera-users] Trouble understanding intersurf results

Elaine Meng meng at cgl.ucsf.edu
Thu Apr 23 08:42:11 PDT 2020 

Hi Kenward,
I sympathize - all my teacher friends are having to generate new materials and methods.

Oh, it sounds like pruning behaves as long as you can use the Intersurf dialog (separate models)!

In that case, if I understand correctly, you can do this in many fewer steps by including 1afv chains K,M in a single model.  I tried splitting 1afv into two models (K,M is one model, everything else is another) using the first command below, but you could also split into any number of components by repeating the "atoms" option:

split atoms :.k,.m

For example, this would split 1afv into 3 parts, the two Fabs and the rest:
split atoms :.k,.m atoms :.h,.l

After using the first split command above, in the Intersurf dialog I chose the two Molecules (actually models):
1afv 1 (#0.1)
1afv 2 (#0.2)

Then I could use Residue centroid distance pruning and Select interface atoms, and it did vary sensibly when I changed the pruning distance.

Something like pruning cutoff 8 and then command "sel up" (same as clicking into graphics window and then pressing up arrow key) to promote selection to whole residues looks pretty decent.

I hope this helps,
Elaine
-----
Elaine C. Meng, Ph.D.                       
UCSF Chimera(X) team
Department of Pharmaceutical Chemistry
University of California, San Francisco

> On Apr 23, 2020, at 12:15 AM, Kenward Vaughan <kay_jay at kvaughan.family> wrote:
> 
> On 4/22/20 10:08 AM, Elaine Meng wrote:
>> Hi Kenward,
>> Sorry for the difficulties.  Intersurf was an implementation based on a paper, and has turned out to be somewhat of a novelty item that doesn't get much use.  I suspect there are bugs, or quirks that I failed to explain in the documentation (likely not having understood it myself).  Its main distinguishing characteristic is that the "inter"surface is not the molecular surface of either set of atoms in the interface but derived from both sets of atoms.  I too am mystified by the behavior of the "prune" option in the command, as it seems that, at least in your example, all prune values (even 0) give the same thing.  I can make a bug report but we are mostly focused on ChimeraX these days.
>> 
> 
> 
> Hi Elaine!
> 
> As is usual, thank you for the detailed reply with the several (detailed) suggestions!   
> 
> My goal is simply to give the students a visualization of the region between the antibody and the epitope itself, hoping that this would give them a better reference when looking at what appears to be a jungle of residues.  Intersurf struck me as the ideal, does it all at once, answer to this idea.  Another for me involves different reps for the two sides (thick line vs. ball and stick, etc.).
> 
> What appears to be working based on your suggestions is a cheap way to get those residues and a usable surface:
> 
> split
> 
> Intersurf B and K, selecting the atoms
> 
> up arrow to collect all atoms of each residue, demo'd by Bonds > show
> 
> name the selection
> 
> redo Intersurf, between B and M, reusing the surface.  Up arrow to gather all atoms of all residues again. Bonds > show.
> 
> Trash the surface in model panel.
> 
> Select mode to append, then select the original named selection.  Now I have all residues for the interface.  New name for the selection.
> 
> Intersect selection with chain B.
> 
> Surf, colorize, transparency. See attached sample...
> 
> 
> 
> I'm unfamiliar with the buried area approach, so for quick results did the above despite the number of steps.  I'm learning all too much with our going online right now.  If there's a fast way to do this by command line I'm all for it, but ugh...
> 
> Don't know how well this'll work with them, but...  Still working on what colors would work well.
> 
> Thank you for the leading suggestions!  
> 
> 
> 
> Kenward
> 
> 
> 
>> Depending on what you want the students to get from the exercise, there may be better alternatives, such as simply selecting atoms or residues with buried area in the interface between the two sets of atoms.  You could color those residues or their molecular surfaces.  (e.g. as in the "open book" image tutorial)
>> 
>> <http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/tutorials/openbook.html>
>> 
>> 
>> Maybe the problem is that the default surface wraps around all the proteins together?  You can get separate ones by using "split" beforehand to make each chain a separate model (as in that tutorial above), or if you can keep the chains in the same model but use "surfcat" to wrap K and M separately from B, e.g. commands:
>> 
>> surfcat bee :.b
>> surfcat kandm :.k:.m
>> surf bee
>> surf kandm
>> 
>> Another command of possible interest is "measure buriedArea" which will assign buriedSESArea and buriedSASArea attribute values to the atoms.  Then one can select atoms by ranges of those values.
>> 
>> <http://www.rbvi.ucsf.edu/chimera/docs/UsersGuide/midas/measure.html#buriedArea>
>> 
>> 
>> measure buriedArea :.b :.k:.m
>> start Render by Attribute 
>>  (...that shows a dialog with histogram of values)
>> select @/buriedSASArea>2.0
>> (... or use the Select tab of that dialog to do the selection... whatever cutoff you like, of course)
>> 
>> Then you can "surf sel" or "color magenta sel" or whatever you were planning to do with the selection of interface atoms.
>>  
>> Sorry about the lame answer re Intersurf.  If there are other questions about how to accomplish various steps of the project, please don't hesitate to ask.  Best regards,
>> Elaine
>> -----
>> Elaine C. Meng, Ph.D.                       
>> UCSF Chimera(X) team
>> Department of Pharmaceutical Chemistry
>> University of California, San Francisco
>> 
>> 
>>> On Apr 21, 2020, at 8:49 PM, Kenward Vaughan <kay_jay at kvaughan.family>
>>>  wrote:
>>> 
>>> On 4/21/20 8:40 PM, Kenward Vaughan wrote:
>>> ...
>>> 
>>> 
>>>> The command I use is
>>>> 
>>>> intersurf   :.b   :.k,.m   pair atom   prune 3   bias 0.5   select
>>>> true
>>>> 
>>>> followed by adjusting the transparency.  The result surprised me,
>>>> both before and after lighting up the atoms and bonds.  I've attached
>>>> a (hopefully small) jpg.
>>>> 
>>>> 
>>> ...
>>> 
>>> By the way, I just realized with further experimentation that a prune distance of 3 gives NO result ("no interface found") using the graphical interface with chain B and either K or M.  Changing this to 10 in my command still gives the same weird results...
>>> 
>>> 
>>> 
>>> 
>>> Kenward
>>> -- 
>>> In a completely rational society, the best of us would aspire to be 
>>> _teachers_ and the rest of us would have to settle for something less, 
>>> because passing civilization along from one generation to the next 
>>> ought to be the highest honor and the highest responsibility anyone 
>>> could have.     - Lee Iacocca
>>> 
>>> Any fool can know. The point is to understand. - Albert Einstein
>>> 
>>> _______________________________________________
>>> Chimera-users mailing list: 
>>> Chimera-users at cgl.ucsf.edu
>>> 
>>> Manage subscription: 
>>> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
> 
> 
> -- 
> In a completely rational society, the best of us would aspire to be 
> _teachers_ and the rest of us would have to settle for something less, 
> because passing civilization along from one generation to the next 
> ought to be the highest honor and the highest responsibility anyone 
> could have.     - Lee Iacocca
> 
> Any fool can know. The point is to understand. - Albert Einstein
> 
> <image.jpg>_______________________________________________
> Chimera-users mailing list: Chimera-users at cgl.ucsf.edu
> Manage subscription: http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users

More information about the Chimera-users mailing list