Check the command you used to load the PDB file and the command you are using to bond. 
You would get that error if you did:
mol=loadpdb structure.pdb

But, then used a different name here:
#BONDING
bond diffName.101.ND2 diffName.508.C1

Or if you are using the incorrect residue numbering. If you check the CPLX.pdb file that gets written out of tleap in my example, you will see what numbering to use, as tleap will renumber your residues starting from 1.

Or if you have the wrong atom names. Basically the error is saying that the atom you are referring to doesn't exist. 

Other things to note:

Oliver

On Thu, Oct 6, 2016 at 12:07 PM, Mahrukh Imtiaz <000005fe14770087-dmarc-reques[log in to unmask]> wrote:

Dear Oliver,
Thanks a lot for your kind response. I have just 3 to 4 missing bonds among branched N-glycans, hence i used the bond command but its not working too. It gives following error:
bond: Argument #1 is type string must be of type: [atom]

Regards,
Mahrukh



On Thursday, October 6, 2016 12:22 PM, Oliver Grant <[log in to unmask]> wrote:


Hi Mahrukh,

If it's just a few bonds it's quickest to do this manually in tleap. I always use tleap (tleap -f leap_inputs.in) so I don't know why you're getting missing parameters in xleap. 
Both xleap and tleap were designed to handle proteins, which are always linear, and thus the program does not handle branching molecules automatically. To get around this you insert a TER card in the PDB where the non-continuous branches start e.g.:

ATOM   5242  O   GLY   507     -12.269 -30.012 -67.337  1.00  0.00
TER
ATOM      2  C1  4YB   508      13.117 -16.557 -28.890  1.00  0.00

the TER is placed there as the 4YB should be connected to an NLN, another e.g.: 

ATOM     73  O6  0MA   513      10.408 -11.034 -18.659  1.00  0.00
TER
ATOM     74  C1  2MA   514      17.112 -11.543 -15.725  1.00  0.00

This is where one branch of an N-linked glycan has ended and the next residue is connected to a residue that is listed earlier in the glycan.
With the TER cards tleap will not automatically try to bond the GLY to the 4YB. Instead you can include bond commands in your tleap input file e.g.:

########### Set Defaults ############################## #####################
set default PBRadii mbondi2
############################## ############################## ################
########### Force Field Inputs ############################## ###############
source /cm/shared/apps/amber14/dat/ leap/cmd/leaprc.GLYCAM_06j-1
source /cm/shared/apps/amber14/dat/ leap/cmd/leaprc.ff14SB
loadAmberParams frcmod.tip5p
############################## ############################## ################

########load Carb################
mol=loadpdb structure.pdb

#BONDING
bond mol.101.ND2 mol.508.C1
bond mol.510.O6 mol.514.C1
etc ...

saveamberparm mol CPLX.prmtop CPLX.rst7
savepdb mol CPLX.pdb

#############Addions########## ####
addIons mol Na+ 0
addIons mol Cl- 0

saveamberparm mol CPLX_Neut.prmtop CPLX_Neut.rst7
savepdb mol CPLX_Neut.pdb

#############Solvate########## ####
solvatebox mol TIP5PBOX 10.0 

saveamberparm mol CPLX_Neut_Sol.prmtop CPLX_Neut_Sol.rst7
savepdb mol CPLX_Neut_Sol.pdb

quit

I've included my entire tleap file for completeness. 

If you have a lot of systems or a lot of N-linked glycans, this is a pain to do. I've semi-automated the process, but it requires you to use some scripts. 

Here is a full link to a how-to: 

If you try those scripts and they don't work for you, please let me know.

Oliver

On Thu, Oct 6, 2016 at 8:36 AM, Mahrukh Imtiaz <000005fe14770087-dmarc- [log in to unmask]> wrote:
Hi everyone,
I am currently trying to perform MD simulations of p-glycoprotein with N-glycans attached to it, but I am having some issues while creating inpcrd and prmtop files via xleap. I have attached glycans via GLYCAM_web but when I load the pdb file into xleap, it shows several missing bonds among the glycan chains including the covalent bond between ASN and GlcNAc. Moreover, when I draw the missing bonds in xleap it gives missing bond, angle and torsion parameter errors. I have been trying to fix this issue but couldn't get any luck. Any help to solve this issue is highly appreciated.

Kind regards,
Mahrukh Imtiaz
COMSATS Institute of Information Technology
Islamabad