Thanks a lot for these precautions, Stephan !
Just to be clear in order that I understand if I do the things correctly.
So to remove the gap I just modelled the missed loop. Then I use the
pdb4amber to fix missed residues and add hydrogens using reduce:
pdb4amber -i ./test.pdb -o ./with_h.pdb -y -p --reduce --no-conect
--add-missing-atoms
From this step, It looks OK but the chain IDs are missing ... Is it
critical for the preparation of the membrane system of the complex (which
is separated by the chain IDS A,B,C in initial pdb) using packmol ?
Many thanks in advance !
Enrico
Il giorno mar 26 mar 2024 alle ore 10:13 Stephan Schott <
s.schott-verdugo.fz-juelich.de> ha scritto:
> Hi,
>
> So, again, it is not packmol removing the TERs, but memembed. I think I
> figured out how to keep them in their code, and this might come to Amber24.
>
> Yes, this would be not ideal if not intended. Still, you should not trust
> blindly the outputs and double check the result of any tool you are using,
> to make sure the structure is indeed what you wanted. pdb4amber, leap,
> packmol-memgen, etc., are helper tools, but relevant aspects like
> protonation states or the structures being parsed could change along the
> way.
>
> Best regards,
>
> Stephan Schott Verdugo
> Biochemist
>
>
> Computational Biophysical Chemistry
> Institut für Bio- und Geowissenschaften / Bioinformatik (IBG-4)
> Forschungszentrum Jülich GmbH
> Wilhelm-Johnen-Straße, 52425 Jülich
> Germany
>
>
> ---------------------------------------------------------------------------------------------
>
> ---------------------------------------------------------------------------------------------
> Forschungszentrum Jülich GmbH
> 52425 Jülich
> Sitz der Gesellschaft: Jülich
> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>
> ---------------------------------------------------------------------------------------------
>
> ---------------------------------------------------------------------------------------------
>
>
> El mar, 26 mar 2024 a la(s) 8:51 a.m., Enrico Martinez (
> jmsstarlight.gmail.com) escribió:
>
>> Thank you very much Stephan !
>>
>> So basically I could be an elegant solution just to fix the packmol in
>> order that it could keep TER for the next tleap processing ...
>>
>> BTW: if I use a dimer protein for packmol (where there is a TER separator
>> between each monomer) does it mean that the packmol will remove the TER and
>> on the next step the TLEAP will create the bond between two monomers
>> embedded into the membrane ? Sound not a great idea at all .. ;-D
>>
>> All the best,
>>
>> Enrico
>>
>> Il giorno lun 25 mar 2024 alle ore 20:04 Stephan Schott <
>> s.schott-verdugo.fz-juelich.de> ha scritto:
>>
>>> Hi,
>>>
>>> The problem doesn't have to do with how the protein is embedded,
>>> memembed or ppm/opm both do a nice job in most cases, is just that memembed
>>> deletes TERs. The --ppm flag will use opm locally, but the option is only
>>> available as a test in versions >23.5. Should be more stable in Amber 24.If
>>> you have a file oriented in the ppm/opm webserver, you should use the
>>> --preoriented flag.
>>>
>>> The N and C-terminal residues are formally different. The H1,H2 and H3
>>> correspond to N-ter residues, but leap "thinks" those are normal residues,
>>> meaning that in the last case, you probably are missing a TER *before*
>>> those residues (the message of the 10 A bond should be a hint). Note that
>>> the file you should be editing if doing this """automatically""" with
>>> packmol-memgen is the packed file, not the input file (so probably
>>> bilayer_XXX.pdb).
>>>
>>> As long as you are aware of what is in your output file, and you are ok
>>> with the gaps, then you can go ahead. You could also consider doing some
>>> overlapping with structures from the AlphaFold DB, if you want to play
>>> around with that.
>>>
>>> Best regards,
>>>
>>> Stephan Schott Verdugo
>>> Biochemist
>>>
>>>
>>> Computational Biophysical Chemistry
>>> Institut für Bio- und Geowissenschaften / Bioinformatik (IBG-4)
>>> Forschungszentrum Jülich GmbH
>>> Wilhelm-Johnen-Straße, 52425 Jülich
>>> Germany
>>>
>>>
>>> ---------------------------------------------------------------------------------------------
>>>
>>> ---------------------------------------------------------------------------------------------
>>> Forschungszentrum Jülich GmbH
>>> 52425 Jülich
>>> Sitz der Gesellschaft: Jülich
>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>
>>> ---------------------------------------------------------------------------------------------
>>>
>>> ---------------------------------------------------------------------------------------------
>>>
>>>
>>> El lun, 25 mar 2024 a la(s) 5:37 p.m., Enrico Martinez (
>>> jmsstarlight.gmail.com) escribió:
>>>
>>>> So.. for the test I took another structure for the same complex but
>>>> again have the same issue with the parametrization in the final step
>>>> (although the protein was embedded correctly):
>>>> /opt/site/ubuntu_22.04/amber/22b/bin/teLeap: Warning!
>>>> There is a bond of 10.448 angstroms between C and N atoms:
>>>> ------- .R<ASP 794>.A<C 11> and .R<ARG 795>.A<N 1>
>>>> FATAL: Atom .R<SER 372>.A<H1 12> does not have a type.
>>>> FATAL: Atom .R<SER 372>.A<H2 13> does not have a type.
>>>> FATAL: Atom .R<SER 372>.A<H3 14> does not have a type.
>>>> FATAL: Atom .R<ASN 420>.A<H1 15> does not have a type.
>>>> FATAL: Atom .R<ASN 420>.A<H2 16> does not have a type.
>>>> FATAL: Atom .R<ASN 420>.A<H3 17> does not have a type.
>>>> FATAL: Atom .R<LYS 462>.A<H1 23> does not have a type.
>>>> FATAL: Atom .R<LYS 462>.A<H2 24> does not have a type.
>>>> FATAL: Atom .R<LYS 462>.A<H3 25> does not have a type.
>>>> FATAL: Atom .R<ASP 477>.A<H1 13> does not have a type.
>>>> FATAL: Atom .R<ASP 477>.A<H2 14> does not have a type.
>>>> FATAL: Atom .R<ASP 477>.A<H3 15> does not have a type.
>>>> FATAL: Atom .R<VAL 511>.A<H1 17> does not have a type.
>>>> FATAL: Atom .R<VAL 511>.A<H2 18> does not have a type.
>>>> FATAL: Atom .R<VAL 511>.A<H3 19> does not have a type.
>>>> FATAL: Atom .R<ARG 795>.A<H1 25> does not have a type.
>>>> FATAL: Atom .R<ARG 795>.A<H2 26> does not have a type.
>>>> FATAL: Atom .R<ARG 795>.A<H3 27> does not have a type.
>>>>
>>>> /opt/site/ubuntu_22.04/amber/22b/bin/teLeap: Fatal Error!
>>>> Failed to generate parameters
>>>>
>>>> Exiting LEaP: Errors = 1; Warnings = 893; Notes = 1.
>>>>
>>>> will it be reasonable just to remove these H1,H2,H3 from indicated
>>>> amino acids ??
>>>>
>>>> I enclose the structure :-)
>>>>
>>>> Thank you !!
>>>>
>>>> Enrico
>>>>
>>>>
>>>> Il giorno lun 25 mar 2024 alle ore 15:37 Enrico Martinez <
>>>> jmsstarlight.gmail.com> ha scritto:
>>>>
>>>>> Well, firstly thank you very much for this information !
>>>>>
>>>>> Actually I thought about the missed TER in the processed pdb ...
>>>>> Should the -ppm work only for the pdb that has been produced by opm
>>>>> web served (with the added dummy atoms for the membrane plane)?
>>>>>
>>>>> Actually in my case I tried it without this option and the protein was
>>>>> perfectly embedded in the bilayer ! And yes I could solve the issue with
>>>>> the manual editing of the structure after it's embedding into the membrane.
>>>>> Finally, here is the original pdb (directly taken from the PDB) with the
>>>>> chains IDs and original TER records, which however also did not parametrize
>>>>> without my manual edition (tleap sent errors due to the hydrogen atoms
>>>>> that are present in this pdb .. )!!
>>>>>
>>>>> Many thanks in advance !!
>>>>>
>>>>> Enrico
>>>>>
>>>>> Il giorno lun 25 mar 2024 alle ore 14:32 Stephan Schott <
>>>>> s.schott-verdugo.fz-juelich.de> ha scritto:
>>>>>
>>>>>> Hi Enrico,
>>>>>>
>>>>>> You didn't tell us what line you are using with packmol-memgen, but I
>>>>>> assume you are aligning it to a membrane slab with the current default,
>>>>>> which is memembed. What I suspect happened there is a bad combi of
>>>>>> successive programs deleting information of the system: pdb4amber when
>>>>>> requested to fill missing atoms, deletes the chain ids, while memembed
>>>>>> seems to delete TER flags. With that, leap at the end is not aware of the
>>>>>> separated chains of the protein. You can either add the chain ids back
>>>>>> before running packmol-memgen, use ppm instead of memembed (added as test
>>>>>> with --ppm in Amber 23.5 or above), or do the manual fix you did by the end
>>>>>> before parametrizing (if using packmol-memgen for getting a suggestion for
>>>>>> the leap input, you can run it first without --parametrize, fix your packed
>>>>>> pdb file, and then add it).
>>>>>>
>>>>>> I will see if I can fix memembed so it doesn't delete TERs to avoid
>>>>>> this.
>>>>>>
>>>>>> Best regards,
>>>>>>
>>>>>> Stephan Schott Verdugo
>>>>>> Biochemist
>>>>>>
>>>>>>
>>>>>> Computational Biophysical Chemistry
>>>>>> Institut für Bio- und Geowissenschaften / Bioinformatik (IBG-4)
>>>>>> Forschungszentrum Jülich GmbH
>>>>>> Wilhelm-Johnen-Straße, 52425 Jülich
>>>>>> Germany
>>>>>>
>>>>>>
>>>>>> ---------------------------------------------------------------------------------------------
>>>>>>
>>>>>> ---------------------------------------------------------------------------------------------
>>>>>> Forschungszentrum Jülich GmbH
>>>>>> 52425 Jülich
>>>>>> Sitz der Gesellschaft: Jülich
>>>>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>>>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>>>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>>>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>>>>
>>>>>> ---------------------------------------------------------------------------------------------
>>>>>>
>>>>>> ---------------------------------------------------------------------------------------------
>>>>>>
>>>>>>
>>>>>> El vie, 22 mar 2024 a la(s) 4:59 p.m., Carlos Simmerling via AMBER (
>>>>>> amber.ambermd.org) escribió:
>>>>>>
>>>>>>> Enrico,
>>>>>>> in the file that you attached, there is a TER between ASN340 and
>>>>>>> GLU15
>>>>>>> (later in the new pdb file these are 338 and 339). because of that,
>>>>>>> pdb4amber will add the missing OXT to 338, and H1/H2/H3 to 339. that
>>>>>>> should
>>>>>>> be fine going into tleap as long as the TER stays there. But tleap
>>>>>>> doesn't
>>>>>>> recognize those as terminal residues, so somehow that TER must have
>>>>>>> been
>>>>>>> lost along the way before you ran tleap.
>>>>>>>
>>>>>>> but then I get confused, because the pdb file output by pdb4amber
>>>>>>> doesn't
>>>>>>> have H1/H2/H3 on SER372 (at least when I ran pdb4amber using your
>>>>>>> command
>>>>>>> on the pdb file that you attached). So again there seems to be things
>>>>>>> changed in the PDB that you end up giving to tleap.
>>>>>>>
>>>>>>> I would suggest focusing on how you go from the pdb output by
>>>>>>> pdb4amber
>>>>>>> into the one that you load into tleap.
>>>>>>> you might also consider trying the prepareforleap function in
>>>>>>> cpptraj.
>>>>>>> building just the protein in gas phase as a simple test might also be
>>>>>>> helpful, then you'll know if the pdb file at each step is set up
>>>>>>> properly.
>>>>>>> then you can do your membrane later.
>>>>>>>
>>>>>>>
>>>>>>> On Fri, Mar 22, 2024 at 11:32 AM Enrico Martinez via AMBER <
>>>>>>> amber.ambermd.org> wrote:
>>>>>>>
>>>>>>> > Dear Amber users !
>>>>>>> >
>>>>>>> > I am working on the preparation of the model for the GPCR protein
>>>>>>> using
>>>>>>> > pack-mol memgen utility of amber tools for creation of the
>>>>>>> membrane and the
>>>>>>> > parametrization of the complex.
>>>>>>> >
>>>>>>> > The initial structure was taken from the pdb, which has some
>>>>>>> missed loops.
>>>>>>> > Since I decided not to restore these missed fragments I removed all
>>>>>>> > hydrogens and then pre-processed the model using
>>>>>>> >
>>>>>>> > pdb4amber -i ./input.pdb -o ./output_ambpro.pdb -y -p
>>>>>>> --add-missing-atoms
>>>>>>> >
>>>>>>> > These re-added missed hydrogens and some of the missed heavy atoms.
>>>>>>> >
>>>>>>> > Then, on the final step of the creation of the membrane complex I
>>>>>>> got the
>>>>>>> > error during the tleap parametrization of the embedded model:
>>>>>>> >
>>>>>>> > There is a bond of 9.221 angstroms between C and N atoms:
>>>>>>> > ------- .R<ASP 798>.A<C 11> and .R<ARG 799>.A<N 1>
>>>>>>> > FATAL: Atom .R<ASN 338>.A<OXT 15> does not have a type.
>>>>>>> > FATAL: Atom .R<GLU 339>.A<H1 16> does not have a type.
>>>>>>> > FATAL: Atom .R<GLU 339>.A<H2 17> does not have a type.
>>>>>>> > FATAL: Atom .R<GLU 339>.A<H3 18> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 372>.A<H1 12> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 372>.A<H2 13> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 372>.A<H3 14> does not have a type.
>>>>>>> > FATAL: Atom .R<ASN 420>.A<H1 15> does not have a type.
>>>>>>> > FATAL: Atom .R<ASN 420>.A<H2 16> does not have a type.
>>>>>>> > FATAL: Atom .R<ASN 420>.A<H3 17> does not have a type.
>>>>>>> > FATAL: Atom .R<LYS 462>.A<H1 23> does not have a type.
>>>>>>> > FATAL: Atom .R<LYS 462>.A<H2 24> does not have a type.
>>>>>>> > FATAL: Atom .R<LYS 462>.A<H3 25> does not have a type.
>>>>>>> > FATAL: Atom .R<ASP 477>.A<H1 13> does not have a type.
>>>>>>> > FATAL: Atom .R<ASP 477>.A<H2 14> does not have a type.
>>>>>>> > FATAL: Atom .R<ASP 477>.A<H3 15> does not have a type.
>>>>>>> > FATAL: Atom .R<VAL 511>.A<H1 17> does not have a type.
>>>>>>> > FATAL: Atom .R<VAL 511>.A<H2 18> does not have a type.
>>>>>>> > FATAL: Atom .R<VAL 511>.A<H3 19> does not have a type.
>>>>>>> > FATAL: Atom .R<LEU 538>.A<OXT 20> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 539>.A<H1 12> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 539>.A<H2 13> does not have a type.
>>>>>>> > FATAL: Atom .R<SER 539>.A<H3 14> does not have a type.
>>>>>>> > FATAL: Atom .R<LYS 595>.A<OXT 23> does not have a type.
>>>>>>> > FATAL: Atom .R<TRP 596>.A<H1 25> does not have a type.
>>>>>>> > FATAL: Atom .R<TRP 596>.A<H2 26> does not have a type.
>>>>>>> > FATAL: Atom .R<TRP 596>.A<H3 27> does not have a type.
>>>>>>> > FATAL: Atom .R<ARG 799>.A<H1 25> does not have a type.
>>>>>>> > FATAL: Atom .R<ARG 799>.A<H2 26> does not have a type.
>>>>>>> > FATAL: Atom .R<ARG 799>.A<H3 27> does not have a type.
>>>>>>> >
>>>>>>> > /opt/site/ubuntu_22.04/amber/22b/bin/teLeap: Fatal Error!
>>>>>>> > Failed to generate parameters
>>>>>>> >
>>>>>>> > Exiting LEaP: Errors = 1; Warnings = 860; Notes = 1.
>>>>>>> >
>>>>>>> >
>>>>>>> > So I carefully examined the structure of the complex created by
>>>>>>> packmol and
>>>>>>> > found that the TER records (in the gaps corresponded to the missing
>>>>>>> > fragments) were removed so tleap did not recognize it correctly.
>>>>>>> For
>>>>>>> > simplicity I added TER records and in other cases (where there is
>>>>>>> no gap) I
>>>>>>> > just removed H1,H2,H3 atoms that were added in the wrong way.
>>>>>>> After this I
>>>>>>> > could parametrize the complex.
>>>>>>> >
>>>>>>> > What could be the best way to overcome such issue ? I enclose the
>>>>>>> initial
>>>>>>> > PDB without hydrogen atoms (before processing by amber-tools) as a
>>>>>>> > template.
>>>>>>> >
>>>>>>> > Many thanks in advance!
>>>>>>> >
>>>>>>> > Enrico
>>>>>>> > _______________________________________________
>>>>>>> > AMBER mailing list
>>>>>>> > AMBER.ambermd.org
>>>>>>> > http://lists.ambermd.org/mailman/listinfo/amber
>>>>>>> >
>>>>>>> _______________________________________________
>>>>>>> AMBER mailing list
>>>>>>> AMBER.ambermd.org
>>>>>>> http://lists.ambermd.org/mailman/listinfo/amber
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>> Forschungszentrum Jülich GmbH
>>>>>>> 52425 Jülich
>>>>>>> Sitz der Gesellschaft: Jülich
>>>>>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>>>>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>>>>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>>>>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>
>>>>>> ------------------------------------------------------------------------------------------------
>>>>>> Forschungszentrum Jülich GmbH
>>>>>> 52425 Jülich
>>>>>> Sitz der Gesellschaft: Jülich
>>>>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>>>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>>>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>>>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>>>>
>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>
>>>>>> ------------------------------------------------------------------------------------------------
>>>>>>
>>>>>
>>>>
>>>>
>>>> ------------------------------------------------------------------------------------------------
>>>>
>>>> ------------------------------------------------------------------------------------------------
>>>> Forschungszentrum Jülich GmbH
>>>> 52425 Jülich
>>>> Sitz der Gesellschaft: Jülich
>>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>>
>>>> ------------------------------------------------------------------------------------------------
>>>>
>>>> ------------------------------------------------------------------------------------------------
>>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------------------------
>>>
>>> ------------------------------------------------------------------------------------------------
>>> Forschungszentrum Jülich GmbH
>>> 52425 Jülich
>>> Sitz der Gesellschaft: Jülich
>>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>>
>>> ------------------------------------------------------------------------------------------------
>>>
>>> ------------------------------------------------------------------------------------------------
>>>
>>
>>
>>
>> ------------------------------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------------------------------
>> Forschungszentrum Jülich GmbH
>> 52425 Jülich
>> Sitz der Gesellschaft: Jülich
>> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
>> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
>> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
>> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>>
>> ------------------------------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------------------------------
>>
>
>
>
> ------------------------------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------------------------
> Forschungszentrum Jülich GmbH
> 52425 Jülich
> Sitz der Gesellschaft: Jülich
> Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498
> Vorsitzender des Aufsichtsrats: MinDir Stefan Müller
> Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende),
> Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens
>
> ------------------------------------------------------------------------------------------------
>
> ------------------------------------------------------------------------------------------------
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Mar 29 2024 - 04:30:01 PDT
