Yes, only a single N-term Met is unlikely to work for Se-Met phasing.
There are many things that can be tried in homology modeling,
including different methods for model generation, use of any cys
residues to improve models, extensive sequence analysis, using hybrid
or multiple models, etc.
Looks like you tried Phase for MR? If so, you should not just look at
tfz but overall Z-score and LLG too to get an idea of the plausibility
of the search model. Even if you can get a low scoring/borderline
solution that is in fact a correct solution but poor score due to low
homology, there are a few things to try in combination to iteratively
rebuild and refine the solution. You can also try out brute force TF
if you think there is a correct RF solution but not getting a correct
TF. You can also look into 6D MR using for a simultaneous rot/trans
search. Can also try some servers to generate multiple models and try
sequentially.
If getting the structure is critical and you have the resources, you
can consider heavy atoms or introducing one or two Leu to Met
mutations at appropriate points and then produce SeMet protein.
I would be happy to take a more detailed look if you want to discuss offline.
Best,
Debanu
On Wed, Jul 6, 2016 at 4:52 PM, Vandna Kukshal <[log in to unmask]> wrote:
> 1 met is the N terminal one so i am worried .
> i tried trimming the loops but did'nt work. i tried homolgy model too.
>
>
>
> On Wed, Jul 6, 2016 at 6:38 PM, Debanu Das <[log in to unmask]> wrote:
>>
>> Hi Vandna,
>>
>> 1 Met out of 115 aa is doable with SeMet especially if you try a
>> multiple sequence alignment and see if the Met is conserved, in which
>> case it is likely to be well-ordered and suitable for phasing.
>>
>> How reliable do you think your MR model is? Did you consider different
>> homology modeling options to create the search model? How about
>> splitting it up into domains or excluding loops from the model while
>> retaining the core (continue to try monomer vs dimer search)? There
>> are many MR strategies to try.
>>
>> Regards,
>> Debanu
>>
>>
>> On Wed, Jul 6, 2016 at 4:28 PM, Vandna Kukshal <[log in to unmask]> wrote:
>> > Thanks Tom and Jacob ,
>> > I am going to try P3 and P6 , This
>> > is
>> > going to ask me to search 12 molecules or more in asymmetric unit as it
>> > is
>> > too big cell.
>> >
>> > On Wed, Jul 6, 2016 at 6:24 PM, Tom Peat <[log in to unmask]> wrote:
>> >>
>> >> Hello Vanda,
>> >>
>> >>
>> >>
>> >> Expanding on the short message from Jacob, you may want to try trigonal
>> >> space groups (P3) as well as the hexagonal (P6) and see if MR may work
>> >> for
>> >> you. This may be due to twinning or it may be that you have a
>> >> noncrystallographic dimer axis which is very close to a
>> >> crystallographic
>> >> axis (or who knows, some other weird effect).
>> >>
>> >>
>> >>
>> >> Cheers, tom
>> >>
>> >>
>> >>
>> >> From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of
>> >> Keller, Jacob
>> >> Sent: Thursday, 7 July 2016 9:19 AM
>> >> To: [log in to unmask]
>> >> Subject: Re: [ccp4bb] MR problem
>> >>
>> >>
>> >>
>> >> Check twinning
>> >>
>> >>
>> >>
>> >> JPK
>> >>
>> >>
>> >>
>> >> From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of
>> >> Vandna Kukshal
>> >> Sent: Wednesday, July 06, 2016 7:14 PM
>> >> To: [log in to unmask]
>> >> Subject: [ccp4bb] MR problem
>> >>
>> >>
>> >>
>> >> Hi,
>> >>
>> >> I am trying to solve one structure with MR and got stucked with
>> >> no
>> >> solution. I have 2.7 A data , i indexed with P3 and Spacegroup what
>> >> pointless gave me is P622. cell dimention is 68 , 68, 348. Data looks
>> >> good
>> >> with R merge ~0.012 and Chi2~1.
>> >>
>> >> Mathew coefficient shows 2-3 molecules in assymetric unit. protein
>> >> exist
>> >> as dimer or as tetramer.
>> >>
>> >> I have model with 28 % identity , i tried MR with all the possible
>> >> spacegroups of P622 but tfz was ~ 5- 6 . I searched for 2-3 monomer
>> >> as
>> >> well as 1 dimer.
>> >>
>> >> in this protein i have only 1 methionine (115 a.a long protein).
>> >>
>> >>
>> >>
>> >> Any suggestion about solving this problem.
>> >>
>> >> thanks
>> >>
>> >>
>> >> .
>> >>
>> >> --
>> >>
>> >> Vandna Kukshal
>> >> Postdoctral Research Associate
>> >> Dept. Biochemistry and Molecular Biophysics
>> >> Washington University School of Medicine
>> >> 660 S. Euclid, Campus Box 8231
>> >> St. Louis, MO 63110
>> >
>> >
>> >
>> >
>> > --
>> > Vandna Kukshal
>> > Postdoctral Research Associate
>> > Dept. Biochemistry and Molecular Biophysics
>> > Washington University School of Medicine
>> > 660 S. Euclid, Campus Box 8231
>> > St. Louis, MO 63110
>
>
>
>
> --
> Vandna Kukshal
> Postdoctral Research Associate
> Dept. Biochemistry and Molecular Biophysics
> Washington University School of Medicine
> 660 S. Euclid, Campus Box 8231
> St. Louis, MO 63110
|