Re: [modeller_usage] Fwd: Modelling of Chimeric protein
To: James Starlight <>
Subject: Re: [modeller_usage] Fwd: Modelling of Chimeric protein
From: flavio seixas <>
Date: Mon, 10 Dec 2012 11:00:20 -0800 (PST)
Cc: Modeller Usage <>
Hi James!
I am glad to help.
The easier way to include hetatoms to model is to treat them as BLK, with the single letter code "."(dot)
In the alignment file, put "." at the position of your hetatms, eg:
aaaaaaaa---------.- (dot refers to hetatm of the first template)
--------bbbbbbbbb-. (dot refers to hetatm of the second template)
ccccccccccccccccc.. (model with hetatm of both templates)
To make your life easier, renumber the templates sequentially. E.g., renumber the first template from 1 to 200 (hetatm is 401) and the second from 201 to 400 (hetatm is 402).
The hetatm of the first and second templates must be renumbered as they will appear in the final model (as alignment above).
Do not forget to put in your script file:
# Read in HETATM records from template PDBs
env.io.hetatm = True
See http://salilab.org/modeller/manual/node18.html for details.
Regards,
Flavio
--- On Mon, 12/10/12, James Starlight <> wrote:
> From: James Starlight <>
> Subject: Re: [modeller_usage] Fwd: Modelling of Chimeric protein
> To: "flavio seixas" <>
> Date: Monday, December 10, 2012, 6:33 PM
> Hi, Flavio!
>
> Thanks for advises! I've used Nedit and exactly the problem
> was in the
> incorrect alignment of y input data. With the corrections
> the modeler
> produced very good models.
>
> Could you also tell me about possible options to include
> HETATM groups
> from templates to the model?
>
> E.g both of my templates have two HETATM groups; from GFP
> protein its
> chromophore group which is covalently bonded to the
> polypeptide chain
> (like retinal in rhodopsin or HEM in globins) of that
> protein.
> from the second globular domain- its small cyclic GMP
> molecule which
> is the diffusive ligand of that protein.
>
> In the final model I'd like to include both of the groups
> from both of
> the templates (chromophore to the GFP part as well as cGMP
> to the
> interior of the second protein).
>
> By the way should I define chromophore group as the part of
> the first
> (gfp) as well as 3rd (output fussed chimera) sequence or
> there is
> another way to its inclusion ?
>
> Thanks for help again,
>
> James
>
> 2012/12/10, flavio seixas <>:
> > Hi James,
> >
> > Modeller results are reproducible. I see you are
> producing just one model.
> > Each time you run your script, it will produce the same
> model with the same
> > error.
> >
> > Try to run more models and compare them. Try at least
> 100. I usually
> > produces 1000 and select the best by modeller dope
> score and procheck
> > evaluation.
> >
> > First of all, try to correct your alignment file.
> >
> > Your alignment file still incorrect (see attached). I
> see that in your file,
> > the sequences are:
> >
> >
> > aaaaaa------*
> > ----bbbbbbb*
> > ccccccccccccccc*
> >
> >
> > All sequences must have the same length!
> >
> > Maybe this is a problem with your text editor. Try to
> use Nedit (for linux).
> > This editor does not have limits for number of columns
> and does not break
> > the lines.
> >
> > I suggest you use Uniprot tab for alignment
> (www.uniprot.org)
> >
> > You can display the alignment in fasta or pir format.
> >
> > regards,
> >
> > Flavio
> >
> >
> >
> >
> > --- On Mon, 12/10/12, James Starlight <>
> wrote:
> >
> >> From: James Starlight <>
> >> Subject: Re: [modeller_usage] Fwd: Modelling of
> Chimeric protein
> >> To: "flavio seixas" <>
> >> Date: Monday, December 10, 2012, 7:26 AM
> >> Flavio,
> >>
> >> below you could find my script as well as new
> alignment
> >> file. That
> >> inpud data produces still wrong result in the
> conformation
> >> of second
> >> template
> >>
> >> # Homology modeling with multiple templates
> >> from modeller import *
> >> # Load standard Modeller classes
> >> from modeller.automodel import * # Load the
> >> automodel class
> >>
> >> log.verbose() # request verbose output
> >> env = environ() # create a new MODELLER
> environment to
> >> build this model in
> >>
> >> # directories for input atom files
> >> env.io.atom_files_directory = ['.',
> '../atom_files']
> >>
> >> a = automodel(env,
> >>
> >> alnfile = 'align-multiple.ali', # alignment
> filename
> >>
> >> knowns = ('1gfl', '3u10'),
> >> # codes of the templates
> >> sequence =
> >> 'seq')
> >> # code of the target
> >> automodel.initial_malign3d= True
> >> a.starting_model= 1
> >> # index of the first model
> >> a.ending_model = 1
> >> # index of the last model
> >>
> >>
> >> # (determines how many models to calculate)
> >> a.make()
> >> # do the
> >> actual homology modeling
> >>
> >>
> >> >P1;1gfl
> >> structureX:1gfl:1 ::238 : :::-1.00:-1.00
> >>
> ASKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKRHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYK---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------*
> >>
> >> >P1;3u10
> >> structureN:3u10:470 :A:671 : :A::-1.00:-1.00
> >>
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DSSRRQYQEKYKQVEQYMSFHKLPADFRQKIHDYYEHRYQGKMFDEDSILGELNGPLREEIVNFNCRKLVASM
> >>
> PLFANADPNFVTAMLTKLKFEVFQPGDYIIREGTIGKKMYFIQHGVVSVLTKGNKEMKLSDGSYFGEICLLTRGRRTASV
> >> RADTYCRLYSLSVDNFNEVLEEYPMMRRAFETVAIDRLDRIGKKNSILL*
> >>
> >> >P1;seq
> >> sequence:seq: : :
> >> : ::: 0.00: 0.00
> >>
> ASKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKRHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYKDSSRRQYQEKYKQVEQYMSFHKLPADFRQKIHDYYEHRYQGKMFDEDSILGELNGPLREEIVNFNCRKLVASMPLFANADPNFVTAMLTKLKFEVFQPGDYIIREGTIGKKMYFIQHGVVSVLTKGNKEMKLSDGSYFGEICLLTRGRRTASVRADTYCRLYSLSVDNFNEVLEEYPMMRRAFETVAIDRLDRIGKKNSILLH*
> >>
> >>
> >> by the way do you know any web servers for
> produccing such
> >> multiple
> >> ali files for modeller input like
> >> aaaa----
> >> ----bbbb
> >>
> >> for protein aaaabbbb
> >> ?
> >>
> >> ( the above file I've done manually)
> >>
> >> James
> >>
> >>
> >> 2012/12/9, flavio seixas <>:
> >> > Hi James.
> >> >
> >> > Maybe if you post your script, it becomes more
> easier
> >> to help.
> >> >
> >> > But, from your post I can see that the lenght
> of your
> >> target and sequence
> >> > alignment are not equal.
> >> >
> >> > See from http://salilab.org/modeller/9.10/manual/node21.html
> >> >
> >> > that the differences in the lenght are filled
> with "-"
> >> >
> >> > Maybe you should check this.
> >> >
> >> > Regards,
> >> >
> >> > Flavio
> >> >
> >> >
> >> > --- On Sun, 12/9/12, James Starlight <>
> >> wrote:
> >> >
> >> >> From: James Starlight <>
> >> >> Subject: Re: [modeller_usage] Fwd:
> Modelling of
> >> Chimeric protein
> >> >> To: "flavio seixas" <>
> >> >> Date: Sunday, December 9, 2012, 6:23 PM
> >> >> Hi Flavio!
> >> >>
> >> >> In the align-multiple.ali file
> >> >> structureN:3u10:470 ::203 :
> :::-1.00:-1.00
> >> >> 470 is the number of first residue (in pdb
> file)
> >> and 203 is
> >> >> the length
> >> >> of the sequence of that protein
> >> >>
> >> >> I have superimpose both of the templates
> and locate
> >> them in
> >> >> the
> >> >> desired location in the initial pdb's
> templates.
> >> >>
> >> >>
> >> >> Is there any string that I should add to
> my script
> >> ?
> >> >> ( that example I have used
> >> >> http://salilab.org/modeller/9.10/manual/node21.html )
> >> >>
> >> >> Also I've added the string
> >> automodel.initial_malign3d= True
> >> >> but there were no difference in the output
> modell
> >> >>
> >> >> James
> >> >>
> >> >> 2012/12/9, flavio seixas <>:
> >> >> > Hi James,
> >> >> >
> >> >> > I have 2 suggestions.
> >> >> >
> >> >> > 1) The alignment file of your second
> template
> >> is set:
> >> >> >> >P1;3u10
> >> >> >> structureN:3u10:470 ::203 :
> >> :::-1.00:-1.00
> >> >> >
> >> >> > I think the correct must be (first
> residie to
> >> last
> >> >> residue):
> >> >> >> >P1;3u10
> >> >> >> structureN:3u10:203 ::470 :
> >> :::-1.00:-1.00
> >> >> >
> >> >> >
> >> >> > 2) Did you put both templates in the
> same
> >> coordinated
> >> >> system? I mean, did
> >> >> > you superpose (Secondary Strucuture
> Match
> >> supermpose)
> >> >> both the templates in
> >> >> > a way that generated models will
> reflect the
> >> template
> >> >> structure?
> >> >> >
> >> >> > Regards
> >> >> >
> >> >> > Flavio
> >> >> >
> >> >> > --- On Sun, 12/9/12, James Starlight
> <>
> >> >> wrote:
> >> >> >
> >> >> >> From: James Starlight <>
> >> >> >> Subject: [modeller_usage] Fwd:
> Modelling
> >> of
> >> >> Chimeric protein
> >> >> >> To: "modeller_usage" <>,
> >> >> "Modeller Caretaker"
> >> >> >> <>
> >> >> >> Date: Sunday, December 9, 2012,
> 11:25 AM
> >> >> >> Dear Modeler Users!
> >> >> >>
> >> >> >>
> >> >> >> Recently I've done my chimeric
> protein
> >> based on
> >> >> two
> >> >> >> different
> >> >> >> templates in accordance to the
> >> >> >> http://salilab.org/modeller/9.10/FAQ.html#1
> >> >> >>
> >> >> >> using that script
> >> >> >> http://salilab.org/modeller/9.10/manual/node21.html
> >> >> >>
> >> >> >> below the alignment of my
> templates as
> >> well as
> >> >> sequence
> >> >> >>
> >> >> >> >P1;1gfl
> >> >> >> structureX:1gfl:1 ::238 :
> >> :::-1.00:-1.00
> >> >> >>
> >> >>
> >>
> ASKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKRHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYK---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------*
> >> >> >>
> >> >> >> >P1;3u10
> >> >> >> structureN:3u10:470 ::203 :
> >> :::-1.00:-1.00
> >> >> >>
> >> >>
> >>
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------DSSRRQYQEKYKQVEQYMSFHKLPADFRQKIHDYYEHRYQGKMFDEDSILGELNGPLREEIVNFNCRKLVASMPLFANADPNFVTAMLTKLKFEVFQPGDYIIREGTIGKKMYFIQHGVVSVLTKGNKEMKLSDGSYFGEICLLTRGRRTASVRADTYCRLYSLSVDNFNEVLEEYPMMRRAFETVAIDRLDRIGKKNSILL.*
> >> >> >>
> >> >> >> >P1;seq
> >> >> >> sequence:seq: : :
> >> >> >> : ::: 0.00: 0.00
> >> >> >>
> >> >>
> >>
> ASKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKRHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYKDSSRRQYQEKYKQVEQYMSFHKLPADFRQKIHDYYEHRYQGKMFDEDSILGELNGPLREEIVNFNCRKLVASMPLFANADPNFVTAMLTKLKFEVFQPGDYIIREGTIGKKMYFIQHGVVSVLTKGNKEMKLSDGSYFGEICLLTRGRRTASVRADTYCRLYSLSVDNFNEVLEEYPMMRRAFETVAIDRLDRIGKKNSILLH*
> >> >> >>
> >> >> >>
> >> >> >> As the result I've obtain model
> where
> >> first
> >> >> template (
> >> >> >> beta-can GFP)
> >> >> >> was in correct form but the
> conformation
> >> of the
> >> >> second
> >> >> >> protein was
> >> >> >> differ from the used template (
> pdb id
> >> 3u10).
> >> >> >>
> >> >> >>
> >> >> >> Could you tell me why my model
> was so
> >> distorted and
> >> >> how I
> >> >> >> can improve
> >> >> >> it further?
> >> >> >>
> >> >> >> Thanks for help
> >> >> >>
> >> >> >> James
> >> >> >>
> >> >> >> 2012/8/8, Modeller Caretaker
> <>:
> >> >> >> > On 8/8/12 6:27 AM, James
> Starlight
> >> wrote:
> >> >> >> >> I want to model chimeric
> protein
> >> which
> >> >> consist of
> >> >> >> two fussed proteins
> >> >> >> >> ( tail to head fussion C
> to N
> >> termi).
> >> >> It's
> >> >> >> important that both of that
> >> >> >> >> proteins have known
> spatial
> >> structures
> >> >> which could
> >> >> >> be used as the
> >> >> >> >> templates.
> >> >> >> >
> >> >> >> > http://salilab.org/modeller/9.10/FAQ.html#1
> >> >> >> >
> >> >> >> >> Is there any way to use
> both of
> >> the
> >> >> templates to
> >> >> >> guide such modelling
> >> >> >> >> ? In the input option
> I've found
> >> only
> >> >> possibility
> >> >> >> to use one template.
> >> >> >> >
> >> >> >> > 'knowns' can be a list of
> multiple
> >> templates.
> >> >> See
> >> >> >> > http://salilab.org/modeller/9.10/manual/node21.html
> >> >> >> >
> >> >> >> > Ben Webb, Modeller
> Caretaker
> >> >> >> > --
> >> >> >> >
> >> >> >> http://www.salilab.org/modeller/
> >> >> >> > Modeller mail list:
> >> >> >> > http://salilab.org/mailman/listinfo/modeller_usage
> >> >> >> >
> >> >> >>
> >> >> >>
> >> _______________________________________________
> >> >> >> modeller_usage mailing list
> >> >> >>
> >> >> >> https://salilab.org/mailman/listinfo/modeller_usage
> >> >> >>
> >> >> >
> >> >>
> >> >
> >>
>