Dear Junmei,
Ok I understand now the problem.
Thank you very much for the explanation and help,
loris...
________________________________
From: Wang, Junmei <JUW79.pitt.edu>
Sent: Thursday, June 1, 2017 2:49:56 PM
To: Loris Moretti; AMBER Mailing List; david.case.rutgers.edu
Subject: Re: [AMBER] amino-cyclopropyl parameters
Dear Loris,
The atom type definition of your molecule is all right. The problem is that the nitrogen in 3 and 4-memberred ring is also recognized as n3/nh/n4/n.
When we did statistics for force field parameters like cx-cx-n3, both types of n3 (in and out of 3 and 4-memberred rings) are considered. However, there are more molecules having cx-cx-n3 (n3 is ring atom) in the data set and the mean value of the bond angle is around 60 degree. I also want to point out that the cx-cx-n3 bond angles where n3 is NOT a ring atom were considered as outliers in our statistical analysis.
The solution is to add new atom types for nitrogen in 3-memberred and 4-memberred rings. I am now working on this issue.
I will also improve parmchk2 to give a warning if the bond lengths/bond angles deviates far away from the references values.
All the best
Junmei
________________________________
From: Loris Moretti <lom.nuevolution.com>
Sent: Thursday, June 1, 2017 3:37:05 AM
To: Wang, Junmei; AMBER Mailing List; david.case.rutgers.edu
Subject: Re: [AMBER] amino-cyclopropyl parameters
Dear Junmei,
I understand that nitrogen in 3 and 4-membered ring needs to have special atom types to define specific bond and angles parameters. I clearly see the need, but just to be sure to be understood correctly, the nitrogen (of the amino group) in the example I gave is substituted to a 3-membered ring (cyclopropane) and not in the ring itself.
C ---- C
\ /
C --- N
I've tried several examples and the distortion was only with the cyclopropyl-n3 and cyclopropyl-nh. The same cases with 4-membered ring were fine.
I hope this helps...thanks,
loris...
________________________________
From: Wang, Junmei <JUW79.pitt.edu>
Sent: Wednesday, May 31, 2017 3:35:34 PM
To: Loris Moretti; AMBER Mailing List; david.case.rutgers.edu
Subject: Re: [AMBER] amino-cyclopropyl parameters
Dear All,
This issue exists for a long time. We didn't have special atom types for n3/n4/nh/n in 3 and 4-memberred rings. The best solution is to introduce bond and ring types into residue topology and force field files, which means the leap and parmchk programs need to change to recognize the bond and ring types. We are now working on this.
During the mean time, new atom types will be introduced to temporarily solve the problem.
ni - n3 in 3-memberred ring
nj - nh in 3-memberred ring
nk - n4 in 3-memberred ring
nl - n in 3-memberred ring
nm - n3 in 4-memberred ring
nn - nh in 4-memberred ring
np - n4 in 4-memberred ring
nq - n in 4-memberred ring
New bond and angle parameters will be updated. I will send out a post once it is done.
All the best
Junmei
________________________________
From: Loris Moretti <lom.nuevolution.com>
Sent: Wednesday, May 31, 2017 3:32:48 AM
To: AMBER Mailing List; david.case.rutgers.edu
Cc: Wang, Junmei
Subject: Re: [AMBER] amino-cyclopropyl parameters
Yes, of course. Here in attachment the pdb I've used, which is the minimal example for the case.
I see what you mean with the cx-cx-na and I've experienced another odd parameter of a similar kind, cx-cx-nh which leads to a wrong minimized structure.
I tested the transposition of values and I get reasonable results.
Thanks,
loris...
________________________________
From: David A Case <david.case.rutgers.edu>
Sent: Tuesday, May 30, 2017 3:52:05 PM
To: AMBER Mailing List
Cc: Junmei Wang
Subject: Re: [AMBER] amino-cyclopropyl parameters
On Tue, May 30, 2017, Loris Moretti wrote:
>
> In the gaff parameters file dat/leap/parm/gaff2.dat the line
> corresponding to the angle term I need is:
>
> cx-cx-n3 114.892 59.590 SOURCE4_SOURCE5 400 0.3281
>
> In the literature one can find that such angle should be around 114
> degrees. So, my guess is that the 2 values have been exchange by
> mistake, in fact once replaced the simulation went fine.
Thanks for reporting this....cc-ing to Junmei for his input. Junmei:
Some of the cx-cx-XX entries look quite odd: e.g. compare cx-cx-n3 to
cx-cx-na. But I suspect this is more of an atom-typing glitch than a
transposition of values. It looks like antechamber is expecting cx-cx-n3
to be in a three-member ring, but that is not happening here (cf. pictures
as
http://archive.ambermd.org/201705/0351.html<
https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Farchive.ambermd.org%2F201705%2F0351.html&data=01%7C01%7Cjuw79%40pitt.edu%7C2c2e781184cb47821cc308d4a7f73ea4%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=xnzPIOojanjRDY7UVXpHD4c4aRhyDBNgBoa1BpM6CwQ%3D&reserved=0>).
Loris: can you post the pdb file you fed to antechamber? We'll try to
get to the bottom of this.
...thanks...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber<
https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.ambermd.org%2Fmailman%2Flistinfo%2Famber&data=01%7C01%7Cjuw79%40pitt.edu%7C2c2e781184cb47821cc308d4a7f73ea4%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=VeFVNl3%2B%2F2zPPZ3TE78c2EjWf4kyRg9wBEvk4wueXP4%3D&reserved=0>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jun 01 2017 - 06:30:04 PDT
