[phenixbb] phenix.find_ncs

Thomas C. Terwilliger terwilliger at lanl.gov
Wed Jan 11 19:29:11 PST 2012


Hi Ursula,

Thanks for sending me the file...you can find all 3 groups this way:

   phenix.simple_ncs_from_pdb your_pdb.pdb  \
       maximize_size_of_groups=true

Here the parameter maximize_size_of_groups is available in simple_ncs_from
pdb (but not from find_ncs):

maximize_size_of_groups= False You can request that the scoring be
                                  set up to maximize the number of members in
                                  NCS groups

In your case the chains in your file are not all quite the same, so the
ncs-finding algorithm took the group that gave the longest match (and
skipped one chain).  Setting maximize_size_of_groups=true will change the
scoring to try and keep as many members as possible.  I think that I will
make this the default...

All the best,
Tom T



>> I just did waht you suggested. I ran phenix.find_ncs on the problem group
>> only. This time it found all 3 chains in the group.
>>
>> Ursula
>>
>> On Wed, Jan 11, 2012 at 10:16 AM, Thomas C. Terwilliger <
>> terwilliger at lanl.gov> wrote:
>>
>>> Hi Ursula,
>>>
>>> I am not sure why find_ncs did not get the 3rd chain for the third
>>> group.
>>> I think that when you run phenix.refine it is (for the moment) actually
>>> just calling find_ncs to get the NCS groups, so the groups found there
>>> should (if everything is working the way I think it should) be identical
>>> to the ones you would find with phenix.find_ncs.
>>>
>>> Here is something to try: Split your model into just protein #3 (you can
>>> use phenix.pdbtools and keep="chain C or chain F or chain I" to do
>>> this),
>>> and also just protein #1 and #2.  Then run find_ncs separately on these.
>>> Can find_ncs get the ncs for your 3 copies of protein #3 all by
>>> themselves?
>>>
>>> If not...then it indicates that perhaps one of the chains is actually
>>> different (maybe very different conformation, maybe mostly missing,
>>> maybe
>>> different sequence...?)
>>>
>>> All the best,
>>> Tom T
>>>
>>>
>>>  >> I am working with a complex of 3 different proteins, and there are 3
>>> >> complexes in the asymmetric unit. During phenix.refine I used
>>> automatic
>>> >> NCS
>>> >> and it worked fine. I now wanted to determine the NCS parameters with
>>> >> phenix.find_ncs. The output file gives me 3 NCS groups, 2 groups with
>>> 3
>>> >> chains, and 1 group with 2 chains.I cannot figure out why one chain
>>> is
>>> not
>>> >> included. I tried to run it with only pdb input, or pdb anf map
>>> >> coefficients, same result. Any suggestions hat else to try?
>>> >>
>>> >> Thanks
>>> >>
>>> >> Ursula
>>> >>
>>> >> --
>>> >> Ursula Schulze-Gahmen, Ph.D.
>>> >> Assistant Researcher
>>> >> UC Berkeley, QB3
>>> >> 356 Stanley Hall #3220
>>> >> Berkeley, CA 94720-3220
>>> >> _______________________________________________
>>> >> phenixbb mailing list
>>> >> phenixbb at phenix-online.org
>>> >> http://phenix-online.org/mailman/listinfo/phenixbb
>>> >>
>>>
>>> _______________________________________________
>>> phenixbb mailing list
>>> phenixbb at phenix-online.org
>>> http://phenix-online.org/mailman/listinfo/phenixbb
>>>
>>
>>
>>
>> --
>> Ursula Schulze-Gahmen, Ph.D.
>> Assistant Researcher
>> UC Berkeley, QB3
>> 356 Stanley Hall #3220
>> Berkeley, CA 94720-3220
>>



More information about the phenixbb mailing list