When building branched oligosaccharides with GLYCAM, you need to have the linear chain you intend to branch from end with the first branched residue so that you can explicitly specify which open valence position you want to add branches to.
With your sequence:
glycan = sequence { ROH 4GB WLB 3VB 0LB }
You aren't specifying which position of WLB to add 3VB+0LB to, and so leap takes a guess. Since the WLB openings are at positions 3 and 4, it takes the first one which is 3, but that is where you then specify to put the 0SA -- so it's actually attaching BOTH 3VB+0LB and 0SA at glycan.3.O3! Leap refuses to print you a prmtop for this structure, but it does produce a PDB file which upon visualization shows this double attachment at O3 to indeed be the case.
Try this:
glycan = sequence { ROH 4GB WLB }
set glycan tail glycan.3.O4
glycan=sequence { glycan 3VB 0LB }
set glycan tail glycan.3.O3
glycan=sequence { glycan 0SA }
Good luck,
Jodi
On Jul 12, 2012, at 2:58 AM, Tomoshi Kameda wrote:
> Dear amber users and developers,
>
>
> When I try to tleap a branched carbohydrate contain sialic acid,
> tleap of AmberTools12 cannot save .top and .crd file with two errors about torsion angle.
>
> ########### input for tleap
> source leaprc.GLYCAM_06h
> glycan = sequence { ROH 4GB WLB 3VB 0LB }
>
> set glycan tail glycan.3.O3
> branch = sequence { glycan 0SA }
>
> saveamberparm branch comp.top comp.crd
> savepdb branch comp.pdb
> quit
> ################
>
>
> ############# (part of) leap.log
> Building topology.
> Building atom parameters.
> Building bond parameters.
> Building angle parameters.
> Building proper torsion parameters.
> ** No torsion terms for H2-Cg-Os-Cy
> ** No torsion terms for Os-Cg-Os-Cy
> Building improper torsion parameters.
> total 5 improper torsions applied
> Building H-Bond parameters.
> Parameter file was not saved.
> Writing pdb file: comp.pdb
> ######################
>
> However, I investigate molecular structure of WLB and 0SA,
> torsinal angles "H2-Cg-Os-Cy" and "Os-Cg-Os-Cy" donot exist.
> Thus, I guess this behaivor is caused by some bugs...
>
>
>
> For your information,
> 1) if using tleap of AmberTools1.5 with newest patch files,
> this error is not obserbed: .top and .crd are generated.
>
>
> 2) if I try not branched carbohydrate with sialic acid, such as
>> branch = sequence { ROH 4GB 4LB 3VB 0SA }
>> saveamberparm branch comp.top comp.crd
>> savepdb branch comp.pdb
>
> tleap work without error evenif AmberTools12.
>
>
> 3)if I try the example of AmberTools12 manual
> (pp.87, describing branched carbohydrates)
>> source leaprc.GLYCAM_06h
>>
>> glycan = sequence { ROH 4YB 4YB VMB } # linear sequence to branch
>> set glycan tail glycan.4.O3 # set attachment point to the O3 in VMB
>> glycan = sequence { glycan 0MA } # add one of the 0MA
>>
>> set glycan tail glycan.4.O6 # set attachment point to the O6 in VMB
>> BR = sequence { glycan 0MA } # add the other 0MA
>>
>>
>> saveamberparm BR comp.top comp.crd
>> savepdb BR comp.pdb
>
> then, tleap works well without error evenif AmberTools12.
>
> Thus,
> 1) suggests this error is AmberTools12 specific.
> 2) 3) suggest this error occurs onlyif molecule shape is branched and contains sialic acid.
>
>
>
> _______________________________________________
> 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 Thu Jul 12 2012 - 05:30:03 PDT
