Re: [AMBER] PDB file carbohydrates naming problem

From: Rui Chen <rchen6.ualberta.ca>
Date: Fri, 23 Aug 2019 22:44:51 -0600

Dear Lachele,

Thank you for your patience and help.

1. I used GLYCAM to generate the glycans first, however, I got severe
steric clash in 2 out of 8 glycans.
2. I went to http://glycam.org/txt, checked the naming of the
carbohydrate. My pdb file is correct ; ) It's very convenient to check the
correct naming format of specific glycans.
3. I run minimization for the glycosylated protein, it worked fine. I think
I am ok.

Thanks again and have a nice weekend.

Best regards,
Rui



On Fri, Aug 23, 2019 at 6:59 AM Lachele Foley <lf.list.gmail.com> wrote:

> 1. The easiest way is to build it using our online GUI here:
> http://glycam.org/cb Another option is to enter DGlcpNAcb1-OME into
> the box here: http://glycam.org/txt (note the http, not https -
> this is one of the smaller reasons that the current version is being
> retired soon.)
>
> 2. I'm guessing you didn't use our online builder to add glycans to
> your protein (http://glycam.org/gp)? If you were using plain leap, I
> would expect some geometric issues. There are currently no formats,
> that we know of or can use, that will encode complex geometric
> information such as one finds with carbohydrates. We are currently
> working on encoding formats that will allow a system such as leap to
> understand such complexities. For now, these need to be set manually
> or by heuristics built in to software like our Glycoprotein Builder
> GUI or CHARMM's.
>
> The force field is independent of the initial geometry. If an initial
> geometry isn't terrible, a force field can be used to minimize the
> system and improve the geometry. But, if the initial geometry is
> really bad, a force field (generally) will not correct the issue.
> This isn't a problem; to the contrary, it is desirable behavior. If,
> in nature, you were to somehow place two atoms so close to each other
> that their cores are very close to each other, the system would be
> unstable. The force field only tries, incompletely, to mimic the
> behavior of nature. The CHARMM gui is doing some geometric
> manipulations for you in the background. With leap, you simply must
> do them yourself.
>
> If leap built you a prmtop and inpcrd, you should be ok. But, run a
> little simulation to be sure, of course.
>
> On Thu, Aug 22, 2019 at 10:21 PM Rui Chen <rchen6.ualberta.ca> wrote:
> >
> > Dear Lachele,
> >
> > 1. Where can I download the library for GlcpNAcb1-OME?
> >
> > 2. I used CHARMM GUI to generate the glycans, now I am running Amber, in
> > the leap input file, I sourced the GLYCAM force field, even though it
> > worked, I think that's not correct. The reason I use CHARMM GUI to
> generate
> > the glycans not GLYCAM is that I tried to use GLYCAM to generate the
> > glycans, I got severe steric clash, CHARMM GUI worked fine and easier. Do
> > you have any suggestion about the force field?
> >
> > Best regards,
> > Rui
> >
> >
> >
> >
> > On Thu, Aug 22, 2019 at 9:45 AM Rui Chen <rchen6.ualberta.ca> wrote:
> >
> > > Thank you for your help. Where can I download the library
> > > for GlcpNAcb1-OME? I couldn't find it.
> > >
> > >
> > >
> > > On Thu, Aug 22, 2019 at 8:24 AM Lachele Foley <lf.list.gmail.com>
> wrote:
> > >
> > >> 1. Probably. The only way to be certain is to download a
> > >> GlcpNAcb1-OME from our site and open them both up in VMD/Chimera/etc
> > >> and compare the atoms to each other and assign names based on that
> > >> comparison. I think you also need to change C to C2N and O to O2N.
> > >>
> > >> 2. Yes.
> > >>
> > >> 3. If all goes according to plan, yes. If this is not what happens,
> > >> then something went wrong. The NAc group in GlcNAc is often a little
> > >> bent in PDB files.
> > >>
> > >> On Wed, Aug 21, 2019 at 10:18 PM Rui Chen <rchen6.ualberta.ca> wrote:
> > >> >
> > >> > No problem.
> > >> > Just want to double check:
> > >> > 1. I renamed N to N2, CT to CME, that's correct.
> > >> > 2. In order to avoid the problem of hydrogen naming, I delete all
> the
> > >> > hydrogens, since leap will generate the hydrogens for me.
> > >> > 3. I don't need to worry about the improper torsions, minimization
> will
> > >> > solve the problem.
> > >> > Thank you.
> > >> >
> > >> > On Wed, Aug 21, 2019 at 7:36 PM Lachele Foley <lf.list.gmail.com>
> > >> wrote:
> > >> >
> > >> > > We are currently working on the inclusion of glycan processing in
> > >> > > PDB4AMBER. At the moment, its capabilities in that regard are a
> > >> > > little limited. Our biggest holdup right now is getting some
> graph
> > >> > > pattern matching to work. The latter will help make the atom
> renaming
> > >> > > robust even if some atoms are missing, etc. If you know anyone
> who
> > >> > > knows a lot about graph comparisons in the boost libraries, send
> them
> > >> > > to us! :-)
> > >> > >
> > >> > > On Wed, Aug 21, 2019 at 7:02 PM Rui Chen <rchen6.ualberta.ca>
> wrote:
> > >> > > >
> > >> > > > Dear Lachele,
> > >> > > >
> > >> > > > I modified the PDB file according to the naming format you sent
> to
> > >> me,
> > >> > > > there is no specific type for N, I guess it should be N2, no
> > >> specific
> > >> > > type
> > >> > > > for CT, I guess it should be CME. I changed them. Now in my
> leap.log
> > >> > > file,
> > >> > > > I only got warning about improper torsion angles.
> > >> > > >
> > >> > > > It seems like it's better to delete all the H before running
> leap.
> > >> Since
> > >> > > > PDB4AMBER will generate H, I removed all the H by VMD and then
> leap.
> > >> > > >
> > >> > > > I not not sure what's the exact purpose of PDB4AMBER, because
> for
> > >> me I
> > >> > > > still need to manually modify everything.
> > >> > > > [image: Screen Shot 2019-08-21 at 4.29.20 PM.png]
> > >> > > >
> > >> > > > Thanks a lot for your help, very helpful.
> > >> > > >
> > >> > > > Best regards,
> > >> > > > Rui
> > >> > > >
> > >> > > > On Tue, Aug 20, 2019 at 11:32 PM Lachele Foley <
> lf.list.gmail.com>
> > >> > > wrote:
> > >> > > >
> > >> > > > > > Thank you for the quick reply. Please let me know where can
> I
> > >> find
> > >> > > the
> > >> > > > > > naming templates.
> > >> > > > >
> > >> > > > > 0YB is in
> > >> > > > >
> > >> > >
> > >>
> http://glycam.org/docs/forcefield/wp-content/uploads/sites/6/2014/03/GLYCAM_06j-1.prep
> > >> > > > > Search the page to find the 0YB entry.
> > >> > > > >
> > >> > > > > NLN for mid-chain positions is in
> > >> > > > >
> > >> > > > >
> > >> > >
> > >>
> http://glycam.org/docs/forcefield/wp-content/uploads/sites/6/2014/02/GLYCAM_amino_06j_12SB.lib
> > >> > > > > Again, use a search to find that entry. I think your atoms
> are
> > >> named
> > >> > > > > right in NLN, but didn't do a careful check. Leap will tell
> you
> > >> all
> > >> > > > > about it if not.
> > >> > > > >
> > >> > > > > Both file formats are a little strange, but should be mostly
> > >> readable.
> > >> > > > > More info on them is here: http://ambermd.org/formats.html
> > >> > > > > The first is a prep file and the second is an off file.
> > >> > > > >
> > >> > > > > > 1. By the way, PDB4AMBER generated the TER cards after each
> > >> NLN, I
> > >> > > can
> > >> > > > > > manually delete them.
> > >> > > > >
> > >> > > > > Ah. Leap will turn that residue and the next into their
> > >> zwitterionic
> > >> > > > > forms if there is a TER card. Best to delete them. You
> might be
> > >> > > > > able to override that with some sort of leap instruction.
> I'm not
> > >> > > > > sure. Let me know if you have trouble after deleting them.
> > >> > > > >
> > >> > > > > > 2. NLN are 0YB are the correct naming. But as you can see
> in my
> > >> leap
> > >> > > > > input
> > >> > > > > > file, I already source the force field specific for glycans.
> > >> Now the
> > >> > > only
> > >> > > > > > problem is the glycan.
> > >> > > > > > [image: Screen Shot 2019-08-20 at 10.23.18 PM.png]
> > >> > > > >
> > >> > > > > The 0YB *residue* is named correctly. It's the atoms that
> aren't.
> > >> > > > > For example, there are no atoms in the template named N, C,
> O, CT,
> > >> > > > > etc. I think NLN's atoms are ok.
> > >> > > > >
> > >> > > > > This is also the issue with the hydrogens. They all have to
> be
> > >> named
> > >> > > > > correctly or leap fails.
> > >> > > > >
> > >> > > > > > 3. I used PDB2PQR to predict the protonation states of the
> > >> protein,
> > >> > > > > PDB2PQR
> > >> > > > > > generates the hydrogens. I manually deleted all the
> hydrogens.
> > >> Does
> > >> > > it
> > >> > > > > > influence the protonation state of the protein?
> > >> > > > >
> > >> > > > > I suppose you're talking about histidine residues? So long as
> > >> you've
> > >> > > > > named them for the protonation state you want (HID, HIE or
> HIP),
> > >> leap
> > >> > > > > should add in the correct number of hydrogens and assign
> > >> reasonable
> > >> > > > > partial charges to them. I didn't find any instances of
> "HIS" in
> > >> your
> > >> > > > > file, but I did find HID and HIE, so I think you're ok.
> > >> > > > >
> > >> > > > >
> > >> > > > >
> > >> > > > > > On Tue, Aug 20, 2019 at 10:07 PM Lachele Foley <
> > >> lf.list.gmail.com>
> > >> > > > > wrote:
> > >> > > > > >
> > >> > > > > > > Hi!
> > >> > > > > > >
> > >> > > > > > > You shouldn't use TER cards after the NLNs unless they are
> > >> > > actually at
> > >> > > > > > > the end of an amino acid chain. They are amino acid
> > >> (modified),
> > >> > > and
> > >> > > > > > > don't follow the glycan rules.
> > >> > > > > > >
> > >> > > > > > > Removing hydrogens isn't necessary, but their names need
> to
> > >> exactly
> > >> > > > > > > match the templates in the force fields. Folks usually
> > >> prefer to
> > >> > > just
> > >> > > > > > > remove them rather than rename them all. If any of their
> > >> positions
> > >> > > > > > > are critical, leave those in and make sure the names are
> > >> right.
> > >> > > > > > > Otherwise, it's easier to delete them. Or rename them
> all if
> > >> you
> > >> > > > > > > prefer to do that.
> > >> > > > > > >
> > >> > > > > > > Your NLN and 0YB atom names need to exactly match the
> > >> templates as
> > >> > > > > > > well. The same hydrogen-naming logic applies, but you
> must
> > >> ensure
> > >> > > > > > > that the heavy-atom names match.
> > >> > > > > > >
> > >> > > > > > > Do you know how to find the lib and prep files where the
> > >> templates
> > >> > > are
> > >> > > > > > > (for naming)?
> > >> > > > > > >
> > >> > > > > > > On Tue, Aug 20, 2019 at 11:49 PM Rui Chen <
> rchen6.ualberta.ca
> > >> >
> > >> > > wrote:
> > >> > > > > > > >
> > >> > > > > > > > By the way, should I remove all the hydrogens before
> using
> > >> > > PDB4AMBER
> > >> > > > > and
> > >> > > > > > > > why? Thank you.
> > >> > > > > > > >
> > >> > > > > > > > On Tue, Aug 20, 2019 at 9:43 PM Rui Chen <
> > >> rchen6.ualberta.ca>
> > >> > > wrote:
> > >> > > > > > > >
> > >> > > > > > > > > Hello,
> > >> > > > > > > > >
> > >> > > > > > > > > I am preparing a PDB file for running LEaP. Since I
> add
> > >> glycans
> > >> > > > > using
> > >> > > > > > > > > CHARMM GUI, I rename the glycans to NLN (N-linkages to
> > >> ASN)
> > >> > > > > according
> > >> > > > > > > to
> > >> > > > > > > > > AMBER manual (LEaP). However, I got error when I run
> > >> leap. I
> > >> > > > > changed
> > >> > > > > > > the
> > >> > > > > > > > > TER cards, but it still didn't work. I also checked
> the
> > >> > > connections
> > >> > > > > > > between
> > >> > > > > > > > > the carbohydrates and the ASN amino acids using VMD, I
> > >> couldn't
> > >> > > > > find
> > >> > > > > > > > > anything wrong. The leap input file, log file, pdb
> file
> > >> > > (mentioned
> > >> > > > > in
> > >> > > > > > > leap
> > >> > > > > > > > > input file) are attached.
> > >> > > > > > > > >
> > >> > > > > > > > > Could you please give me some clue about how to solve
> the
> > >> > > problem?
> > >> > > > > > > Looking
> > >> > > > > > > > > forward to hearing from you.
> > >> > > > > > > > >
> > >> > > > > > > > > Sincerely,
> > >> > > > > > > > > Rui
> > >> > > > > > > > >
> > >> > > > > > > > _______________________________________________
> > >> > > > > > > > AMBER mailing list
> > >> > > > > > > > AMBER.ambermd.org
> > >> > > > > > > > http://lists.ambermd.org/mailman/listinfo/amber
> > >> > > > > > >
> > >> > > > > > >
> > >> > > > > > >
> > >> > > > > > > --
> > >> > > > > > > :-) Lachele
> > >> > > > > > > Lachele Foley
> > >> > > > > > > CCRC/UGA
> > >> > > > > > > Athens, GA USA
> > >> > > > > > >
> > >> > > > > > > _______________________________________________
> > >> > > > > > > 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
> > >> > > > >
> > >> > > > >
> > >> > > > >
> > >> > > > > --
> > >> > > > > :-) Lachele
> > >> > > > > Lachele Foley
> > >> > > > > CCRC/UGA
> > >> > > > > Athens, GA USA
> > >> > > > >
> > >> > > > > _______________________________________________
> > >> > > > > 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
> > >> > >
> > >> > >
> > >> > >
> > >> > > --
> > >> > > :-) Lachele
> > >> > > Lachele Foley
> > >> > > CCRC/UGA
> > >> > > Athens, GA USA
> > >> > >
> > >> > > _______________________________________________
> > >> > > 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
> > >>
> > >>
> > >>
> > >> --
> > >> :-) Lachele
> > >> Lachele Foley
> > >> CCRC/UGA
> > >> Athens, GA USA
> > >>
> > >> _______________________________________________
> > >> 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
>
>
>
> --
> :-) Lachele
> Lachele Foley
> CCRC/UGA
> Athens, GA USA
>
> _______________________________________________
> 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
Received on Fri Aug 23 2019 - 22:00:02 PDT
Custom Search