[gmx-users] isopeptide bond

Sun Mar 13 11:58:35 CET 2011

On 13/03/2011 9:53 PM, Yulian Gavrilov wrote:
>
> Thank you!
>
> I use the correct “O” in Gly according to .rtp and I checked it with 
> vmd. There is really a new isopeptide bond. When there is no bond, 
> after minimization and equilibration, Gly and Lys just close to each 
> other but they are not connected (in vmd). In my case they are 
> connected (in vmd, pymol).
>

None of these visualization programs read the topology in your .top 
file. They just make guesses based on the geometry of the atoms in the 
coordinate file. Anything you see that it guessed is irrelevant. Read 
your pdb2gmx output, and go and look at the topology to see what atoms 
have made a bond.

Mark

> When I look on step...pdb, one these residues is exploded (it's atoms 
> are far from each other outside of the water box).
>
>
>
> 2011/3/13 Mark Abraham <Mark.Abraham at anu.edu.au 
> <mailto:Mark.Abraham at anu.edu.au>>
>
>     On 13/03/2011 8:55 PM, Yulian Gavrilov wrote:
>>
>>     Dear gmx users,
>>
>>     I just started with gromacs.
>>
>>     Can you help me to find my mistake? I already asked about it, but
>>     I did not understand what to do exactly in my case.
>>
>>     I try to run to add a new *isopeptide bone* to connect Lys and
>>     Gly (to make a tetramer of *ubiquitin*). I use AMBER99 force
>>     field, Gromacs 4.0.5.
>>
>>     What I did:
>>
>>       1.
>>
>>           I changed names of residues according to AMBER rules (LYS
>>           to LYN etc.).
>>
>>       2.
>>
>>           Added new type of residues to ffamber.rtp (LYN -> LYQ and
>>           GLY -> GLQ that are making a new isopeptide bond) and added
>>           a new line to specbond.dat (LYN NZ 1 GLY C 1 0.13 LYQ GLQ)
>>           to make such a bond.
>>
>
>     IIRC, this creates a bond between the lysine side-chain amine N
>     and glycine *backbone* carbonyl C. You must use the atom name for
>     the side-chain carbonyl carbon (see .rtp entry for GLY). If you've
>     done this wrong, then specbond will probably not have made a bond,
>     because the backbone carbonyl C was too far away. You should check
>     your pdb2gmx output carefully.
>
>
>>       1.
>>
>>
>>
>>       2.
>>
>>           Added new bond type, angle type and dihedral type to
>>           ffamber99bon.itp
>>
>>
>>     After running of MD (I've got good minimization and equilibration
>>     – nvt and npt) I've got such an error:
>>
>>
>>     starting mdrun 'Protein in water'
>>
>>     600000 steps, 1200.0 ps.
>>
>>     step 94100, will finish Sun Mar 13 08:15:56 2011
>>
>>     Step 94124, time 188.248 (ps) LINCS WARNING
>>
>>     relative constraint deviation after LINCS:
>>
>>     rms 0.000796, max 0.032792 (between atoms 2454 and 2456)
>>
>>     bonds that rotated more than 30 degrees:
>>
>>     atom 1 atom 2 angle previous, current, constraint length
>>
>>     2454 2457 35.6 0.1522 0.1545 0.1522
>>
>>
>>     And after it the same type of errors with another atoms:
>>
>>     2454 2456 36.1 0.1106 0.1113 0.1090 --> *Gly CA and HA1, HA2*
>>
>>     2454 2455 40.5 0.1114 0.1103 0.1090
>>
>>     .....
>>
>>     2454 2456 90.0 0.1078 0.1479 0.1090
>>
>>     771 773 48.5 0.1011 0.1012 0.1010
>>
>>     ......
>>
>>     771 773 39.8 0.1012 0.1005 0.1010 --> *Gly NZ and HZ1, HZ2*
>>
>>     771 772 34.9 0.1012 0.1030 0.1010
>>
>>     .......
>>
>>     2454 2457 102.1 0.1490 38312886396780544.0000 0.1522
>>
>>     2454 2456 83.0 5.9313 39290317874135040.0000 0.1090
>>
>>     .......
>>
>>     First errors are with atoms from the residues with *new
>>     isopeptide bond*. I suppose, that this bond is not good.
>>
>
>     Seems like a reasonable hypothesis - but do look at 2454 and 2456
>     as well. You have to get out your trajectory and a visualization
>     package and see what is actually going wrong. The warnings can be
>     symptomatic of a problem that started elsewhere.
>
>     You say you've added new interaction types, but I see no reason
>     why you would need to. It's chemically so similar to a backbone
>     peptide that you may as well keep things the same and model it the
>     same way. Regardless, you should probably check that the
>     specbond.dat mechanism has created interactions that make sense.
>     Compare with a normal peptide bond.
>
>     Mark
>
>
>>     Please can you advice me how yo make this isopeptide bond good?
>>
>>     Can I just remove this hydrogens?
>>
>>
>>     -- 
>>
>>     Sincerely,
>>
>>     Yulian Gavrilov
>>
>>
>
>
>     --
>     gmx-users mailing list gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://lists.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at
>     http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>     Please don't post (un)subscribe requests to the list. Use the
>     www interface or send it to gmx-users-request at gromacs.org
>     <mailto:gmx-users-request at gromacs.org>.
>     Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
>
>
>
> -- 
>
> Sincerely,
>
> Yulian Gavrilov
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110313/491cdd8d/attachment.html>