[buster-discuss] Advice on ncs and target restraints

[Oliver Smart]

On Sat, 13 Dec 2014, esserlo at helix.nih.gov wrote:

> Hi,
>
>  this is a question of a buster/tnt novice ...
>
>  I am having a little bit of trouble understanding which residues are
> restrained by NCS and independently of NCS, which residues are restrained by
> a target that is provided.
>
> I am running 'refine' so far in 'auto' mode.
>
> refine -m data -p model  -target ./Reference/combi.pdb -autoncs
>
> I can see in the LIST.html file in each Cycle-## directory with a link to
> auto_gelly_cards.txt the following:
>
> NOTE BUSTER_TARGET01 Reference/combi.pdb
> NOTE BUSTER_SET target01set = All
> NOTE BUSTER_SIM_DEFINE target01 target01set TARGET01_*
> NOTE BUSTER_SIM_RESTRAIN_LSSR target01
>
> Is this all information about the target restraints ?

No

> Is there a way to get
> a list that explicitly states which model residue is being restrained by which
> target residue (as in phenix.refine) ?

Look further down LIST.html you will find a section

   | Interpreting card: NOTE BUSTER_SIM_DEFINE target01 target01set TARGET01_*
   |     NCS restraint to conformation from target file #01
   |     Number of atoms in template set=  2391
   |     Number of related chains (other than template) in SIM set=    1
   |     List of chains to be restrained (excluding template)=   *
   |     Pairwise RMSD, XYZ in Angstroms, for SIM group: target01
   |                 A         01*
   |     A         -----     -----
   |     01*       0.584     -----
   |     Overall RMSD,XYZ for the SIM group is     0.584 Angstroms
   |     Pairwise rms deviation of [individual B's minus their set/chain average] in Angs**2
   |                 A         01*       Bavg
   |     A         -----     -----     43.14
   |     01*       10.01     -----     22.03
   |     Overall RMSD,B   for the SIM group is    10.010 Angs**2
   |     The number of common atoms for the above fits:
   |                 A         01*
   |     A         -----     -----
   |     01*        2357     -----
   |     (compared to the template set that has    2391 atoms in it)
   |     List of atoms that are absent in one or more chains for this group:
   |                                    A    01*
   |                    N       161     +    -
   |                    CA      161     +    -
   |                    C       161     +    -


(I have added "  |" to the start of each line").

The "number of common atoms" details the number of atoms subject to 
restraints. Atoms that are not restrained are listed in the section

"List of atoms that are absent in one "


> The reason why I ask is that it appears that some restraints for some domains
> are not properly applied.
> My target "combi.pdb" contains a number of breaks where I know that the
> reference is not useful. Could this be a problem ?

The requirement is that chain id and residue numbers are the same for
equivalent residues in the structure being refined and the target. See

https://www.globalphasing.com/buster/manual/gelly/manual/gelly4A.html#target

Removing residues that you know are different from the target pdb file is
a sensible thing to do.

>
> Now on to NCS
>
> Does
> NOTE BUSTER_SET ncsautoXcld = Water
> NOTE BUSTER_SET ncsautoXcld = ncsautoXcld + { A|45 A|48 A|53 A|67 A|70 A|73 }
> ....
>
>    ...mean that water and the individually listed residue are excluded from
> any NCS consideration?

Yes this is exactly what is meant.


>  Or are those ranges of residues that are excluded?

No the card above does not specify any residue ranges. BUSTER uses TNT 
type syntax for ranges. So is you wanted to alter the autoncs to exclude
residues in the range chain A 45 to 53 plus chain A 70 to 73 then start 
from the automatically produce cards and edit the file:

NOTE BUSTER_SET ncsMyXcld = Water
NOTE BUSTER_SET ncsMyXcld = ncsMyXcld +  { A|45 - A|53 }
NOTE BUSTER_SET ncsMyXcld = ncsMyXcld +  { A|70 - A|73 }

>
> As much as autoncs is convenient for new users, do experts prefer to define
> their own ncs sets still ?
>

Most people use -autoncs but on occasion, when worryied about detailed 
effects, people manually adapt auto_gelly_cards.txt, see
https://www.globalphasing.com/buster/manual/gelly/manual/gelly4A.html#manual


Hope this helps,

Oliver for buster-develop

--------------------------------------------------------
Oliver Smart http://www.globalphasing.com/people/osmart/