Hi Oliver!

Thank you for the detailed response! I apologize for the late reply; it's been a rather busy week!

I made the changes that you had mentioned and reformatted the carbohydrate.

I'm still getting an odd error with GROMACS when using the converted topology/coordinate files. Below is how I have formatted my Glycan. Attached is my tleap output.

I'm still unsure if this is a tleap issue (input/output attached) or a conversion issue.

I tried using the perl script glycam2gmx.pl, but it outputs an error (below) that upon searching correlates with a need for the software ptraj, which is no longer included in current versions of ambertools. I assume it needs this software to be able to

The ACPYPE tool I was using also included a fix for negative dihedrals force constants for Glycans: https://github.com/austenb28/ACPYPE_FIX

This tool is able to convert the top/crd files, but they fail during early steps of MD preparation.

If it's not a tleap issue, then I think I might skip the conversion and continue with Amber.


Glycam2gmx.pl output

sh: rdparm: command not found
Parsing PDB file lambda.pdb...

boxParams; X: -2000 Y: -2000 Z: -2000
boxAngles; X: 0 Y: 0 Z: 0

    0 residues
    0 sodium ions
    0 chloride ions
    0 waters

Use of uninitialized value $ntotalres in printf at ./glycam2gmx.pl line 311.
    0 TOTAL residues read


Glycan Format:

HETATM 2190  O1  ROH A 452      17.463  64.844  45.461  1.00 17.62           O  
HETATM 2189  C1  2LB A 452      17.907  63.700  46.176  1.00 17.49           C     
HETATM 2191  C2  2LB A 452      17.023  62.505  45.843  1.00 16.53           C
HETATM 2193  C3  2LB A 452      17.583  61.243  46.516  1.00 17.53           C  
HETATM 2194  O3  2LB A 452      16.850  60.099  46.104  1.00 19.22           O  
HETATM 2195  C4  2LB A 452      19.059  61.053  46.160  1.00 16.95           C  
HETATM 2196  O4  2LB A 452      19.174  60.798  44.767  1.00 16.23           O  
HETATM 2197  C5  2LB A 452      19.819  62.324  46.521  1.00 16.94           C  
HETATM 2198  O5  2LB A 452      19.272  63.431  45.796  1.00 16.08           O  
HETATM 2199  C6  2LB A 452      21.295  62.261  46.190  1.00 17.28           C
HETATM 2200  O6  2LB A 452      21.974  63.400  46.704  1.00 18.82           O
HETATM 2192  O2  2LB A 452      15.687  62.763  46.319  1.00 17.71           O 
HETATM 2179  C1  0fA A 453      14.654  62.317  45.491  1.00 16.96           C  
HETATM 2180  C2  0fA A 453      13.301  62.593  46.168  1.00 17.65           C  
HETATM 2181  C3  0fA A 453      13.030  64.102  46.218  1.00 18.33           C  
HETATM 2182  C4  0fA A 453      13.088  64.679  44.808  1.00 18.96           C  
HETATM 2183  C5  0fA A 453      14.463  64.362  44.202  1.00 18.14           C  
HETATM 2184  C6  0fA A 453      14.599  64.833  42.775  1.00 19.90           C  
HETATM 2185  O2  0fA A 453      13.321  62.070  47.482  1.00 19.87           O  
HETATM 2186  O3  0fA A 453      11.751  64.343  46.784  1.00 20.02           O 


Thank you again for helping me!

Kind regards,

Aarya Venkat

Graduate Student

University of Georgia

From: Users of GLYCAM & GLYCAM-Web <[log in to unmask]> on behalf of Oliver Grant <[log in to unmask]>
Sent: Wednesday, October 31, 2018 9:57 AM
To: [log in to unmask]
Subject: Re: Gromacs Glycam tleap issue
Hi again,

I see you have a co-complex, so building the carbohydrate on glycam.org won't help you. Nor will building from sequence. You'll need to reformat the carbohydrate in 1LZJ in GLYCAM nomenclature so that tleap can recognise it and bond it correctly. Try reorganizing the atoms so the autobonding works. Send all the output from tleap if not.


On Wed, Oct 31, 2018 at 2:52 PM Oliver Grant <[log in to unmask]> wrote:
Hi Aarya,

Don't use acpype, unless they have explicitly said that they've updated it to work with GLYCAM. Some terms won't get converted.

You can build the structure on glycam.org/cb and click "download all structures". You'll get the parameter and rst files in the zip file.

The sequence command in leap is for building from prep files. You could do it that way if you want. But note it should be: { ROH 2LB 0fA }. 1fA doesn't make sense there.  More detailed examples are the amber manual. 

If you load try2.pdb into leap, the issue might be that leap will bond the last atom in a residue to the first atom in the next residue, so order the atoms in the PDB file according to how you want the residues to be auto-bonded by leap (i.e. put the O2 as the last entry for residue 2LB), or put TER cards and use the "bond" command (useful for branched carbohydrates). I'd need to see your leap output to help further.



On Wed, Oct 31, 2018 at 2:37 PM Aarya Venkat <[log in to unmask]> wrote:

Hi everyone!

I am attempting to run an MD in GROMACS on a co-crystallized Glycosyltransferase containing a carbohydrate ligand bound.

The ligand is a-L-Fucp-(1-2)-B-D-Galp.

My structure is 1LZJ is from: https://www.rcsb.org/structure/1LZJ

To parametrize this ligand, I am using a combination of tleap and either acpype or Glycam2gmx.pl to convert the amber parm/crd outputs into gromacs format, but neither are giving me working input files (fails at generating ions due to topology incompatibility/Glycam2gmx output finds no residues, where acpype does find residues).

I have attached my ligand/PDB as well as crd and top files from tleap.

I am thinking this may be due to my tleap file. While it outputs the files without error, I am concerned that I am perhaps not including something here.

My code is simply:

source leaprc.GLYCAM_06j-1
loadamberprep GLYCAM_06j-1.prep
mol = loadpdb try2.pdb
saveamberparm mol nacc.parm7 nacc.rst7

I have seen others build sequences from their tleap.

ie: seq1 = sequence { ROH 2LB 1fA }

I wasn't sure if I had to do this given that I loaded the pdb in. Please let me know if you need any other information. I would appreciate any advice you have. Thank you!

Kind regards,

Aarya Venkat