[chimerax-users] Dali server alignment

Mon May 2 01:43:13 PDT 2022

Dear Elaine,

Part (A) sounds great! I didn't realise we could access these matrices  
so easily.

For (B), there is a list of structural equivalences that you could  
probably use.

In your example of 2gbp-A and 2fw0-A, there is just one entry:

# Structural equivalences
    1: 2gbp-A 2fw0-A     2 - 306 <=>    1 - 305   (ASP    2  - PHE   
306  <=> ASP    2  - PHE  306 )

For a more complicated case that I made with 3h5v-A and 3qel-A  
(http://ekhidna2.biocenter.helsinki.fi/barcosel/tmp//ESTfGn3xaNs//3h5vA.txt),  
we get many more:

# Structural equivalences
    1: 3h5v-A 3qel-A     1 -  11 <=>    2 -  12   (SER    3  - PRO    
13  <=> PRO   24  - SER   34 )
    1: 3h5v-A 3qel-A    18 -  25 <=>   16 -  23   (TYR   20  - MET    
27  <=> HIS   38  - ALA   45 )
    1: 3h5v-A 3qel-A    26 -  32 <=>   25 -  31   (VAL   28  - GLU    
34  <=> ASN   47  - HIS   53 )
.
.
.

    1: 3h5v-A 3qel-A   371 - 376 <=>  342 - 347   (LYS  373  - LEU   
378  <=> TYR  392  - ASP  397 )

ProDy can parse something similar for searches against the whole PDB,  
but I don't know if the format is similar enough for that to be  
helpful. Maybe I can let you know later.

Best wishes
James

Elaine Meng via ChimeraX-users <chimerax-users at cgl.ucsf.edu> escribió:

> Dear Hong,
> I broke this down into (A) showing Dali 3D superposition, and (B)  
> showing sequence alignment and RMSD values.
>
> (A) showing Dali 3D superposition.  Unfortunately the Dali Server  
> <http://ekhidna2.biocenter.helsinki.fi/dali/> does not give anything  
> as convenient as a PDB file!
>
> However, you can use the transformation matrix provided by the Dali  
> Server with the ChimeraX "view matrix" command to superimpose the  
> structures.  You would not use matchmaker, because Dali already  
> calculated the superposition.  You simply want to apply the Dali  
> superposition.
>
> Details:
>
> (1) I tried the Dali Server with first structure 2gbpA and second  
> structure 2fw0A.  Then I got a results page that looks like this:
>
> (2) then I clicked the "Parseable (txt)" link and I got a page that  
> looks like this.  (You could also save it as a text file from your  
> web browser).
>
>
> (3) then I reformatted those 12 numbers in the matrix to use in the  
> ChimeraX "view matrix" command.  In ChimeraX, I opened 2gbp (model  
> #1) and 2fw0 (model #2) and then used the following command, which  
> should all be one line even though the email may display as multiple  
> lines:
>
> view matrix mod  
> #2,-0.464817,-0.320176,0.825489,65.924568,-0.856265,0.399781,-0.327087,21.951881,-0.225289,-0.858873,-0.459981,18.394049
>
> See help
> <https://rbvi.ucsf.edu/chimerax/docs/user/commands/view.html#matrix>
>
> (B) Showing sequence alignment and RMSD values.
>
>  I couldn't figure out how to get a sequence alignment file that can  
> be used as ChimeraX input from the Dali server either!  But it  
> sounds like you were able to do it.  If so, you can just open the  
> file in ChimeraX and show the RMSD header and use it for coloring:
>
> sequence header rmsd show
> color byattribute seq_rmsd #2
>
> You need to show the RMSD header in order to be able to color by the  
> seq_rmsd header values.
> <https://rbvi.ucsf.edu/chimerax/docs/user/commands/sequence.html#header>
>
> You can use different coloring if you like:
> <https://rbvi.ucsf.edu/chimerax/docs/user/commands/color.html#byattribute>
> <https://rbvi.ucsf.edu/chimerax/docs/user/commands/color.html#palette-options>
>
> Since I couldn't figure out how to get the sequence alignment from  
> Dali, I used ChimeraX for that part too.  If it is just a pairwise  
> comparison, actually you can use ChimeraX matchmaker to show the  
> sequence alignment withRMSD header:
>
> mm #2 to #1 show true
>
> But *after* that matchmaker command, you would then need to use the  
> "view matrix" command from part (A) above, to apply the Dali  
> superposition instead of simply using the matchmaker superposition,  
> if the Dali one is what you want.  Then you can color by the  
> seq_rmsd attribute as shown above.
>
> If it is >2 structures, then you have to resort to other methods of  
> getting the sequence alignment.  One way is to open all of the  
> structures in ChimeraX, then use the "seq align" command to use a  
> web service to make a multiple sequence alignment.  This command is  
> a recent feature, so it is not in the version 1.3 release -- you  
> would need to get a new daily build to use it.  Example if you want  
> sequence alignment of #1 chain A, #2 chain A, #3 chain B:
>
> seq align #1/A #2/A #3/B
> sequence header rmsd show
>
> See "seq align" help:
> <https://rbvi.ucsf.edu/chimerax/docs/user/commands/sequence.html#align>
>
> 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 29, 2022, at 8:14 PM, HONG ZHAN via ChimeraX-users  
>> <chimerax-users at cgl.ucsf.edu> wrote:
>>
>> Dear all,
>> I am trying to align molecules based on Dali server output.  
>> However, I can’t download the superposition results from the Dali  
>> server webpage (or anyone know?), but I can generate an alignment  
>> file with 2D structural alignment from the Dali server. Is there a  
>> way to use matchmaker to superpose molecules based on output  
>> alignment from Dali server? How to run structure conservation  
>> calculation to color the molecule? If I just give Dali’s alignment  
>> file, it still calculate conservation based on sequence not  
>> structure.
>> Best,
>> Hong