Re: [NAMD] Question about patches

From: JC Gumbart (gumbart_at_physics.gatech.edu)
Date: Tue Dec 26 2017 - 09:26:12 CST

Good point! But there is a way around this as well, which should always be run when building a new PSF, anyway. Just add the line

regenerate angles dihedrals

to your psfgen script after building all segments and applying patches. I believe angles and dihedrals do get rebuilt when making segments but not when applying patches.

Best,
JC

> On Dec 26, 2017, at 4:43 AM, Lennart Nilsson <Lennart.Nilsson_at_ki.se> wrote:
>
> For completeness: <>
> You first have to delete atom H61, which in the canonical ADE residue is bonded to N6; if you don’t do this, the bonds, angles and dihedrals that are associated with H61 will be kept.
> PRES M6NA 0.0
> DELETE ATOM H61
> .
> .
>
> You should also check that you have the expected total charge and number of atoms, bonds, angles, impropers and dihedrals when you apply the patch. Not always easy, but in this case you can compare with what you get with RESI 6MA – and you can even compare energies.
>
> Best wishes,
> Lennart Nilsson
>
> From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On Behalf Of The Cromicus Productions
> Sent: den 26 december 2017 01:05
> To: The Cromicus Productions <thecromicusproductions_at_gmail.com>
> Cc: namd-l_at_ks.uiuc.edu
> Subject: Re: namd-l: [NAMD] Question about patches
>
> I figured it out! This happened because I forgot to add the files
> topology top_all36_na.rtf
> topology top_all36_cgenff.rtf
> topology top_all36_carb.rtf
>
> Now it's working, thanks everyone and Merry Christmas!
>
> On Mon, Dec 25, 2017 at 6:53 PM, The Cromicus Productions <thecromicusproductions_at_gmail.com <mailto:thecromicusproductions_at_gmail.com>> wrote:
> I tried with "mutate" and looks like it ALMOST works, as I get the following error:
> psfgen) mutating residue 4 from ADE to 6MA
> psfgen) Info: generating structure...psfgen) unknown atom type HGA3
> psfgen) add atom failed in residue 6MA:1
> failed!
> ERROR: failed on end of segment
>
> On Mon, Dec 25, 2017 at 6:29 PM, The Cromicus Productions <thecromicusproductions_at_gmail.com <mailto:thecromicusproductions_at_gmail.com>> wrote:
> Thanks for the replies! I know that it is defined as a residue, however, I'm not very familiar with the force field files. My question was, is it possible/easy to
> define a patch that converts an A residue into a 6MA? If so, how can I do that/learn to do that?
>
> I tried to add the following but get an error "psfgen) applying patch M6NA to 1 residues
> psfgen) unknown atom type HGA3
> psfgen) add atom failed in patch M6NA
> ERROR: failed to apply patch"
>
> PRES M6NA 0.00 ! patch for converting A into N6MA
> ! use in PATCH statement followed by AUTOgenerate ANGLes DIHE
> GROUP
> ATOM N9 NG2R51 -0.01 ! H61
> ATOM C8 CG2R53 0.43 ! \
> ATOM H8 HGR52 0.08 ! H62-CM6 H6
> ATOM N7 NG2R50 -0.85 ! / \ /
> ATOM C5 CG2RC0 0.32 ! H63 N6
> ATOM C6 CG2R64 0.45 ! |
> ATOM N6 NG311 -0.45 ! C6
> ATOM H6 HGPAM1 0.33 ! // \
> ATOM N1 NG2R62 -0.78 ! N1 C5--N7\\
> ATOM C2 CG2R64 0.49 ! | || C8-H8
> ATOM H2 HGR62 0.14 ! C2 C4--N9/
> ATOM N3 NG2R62 -0.86 ! / \\ / \
> ATOM C4 CG2RC0 0.52 ! H2 N3 \
> ATOM CM6 CG331 -0.08 ! \
> ATOM H61 HGA3 0.09 ! \
> ATOM H62 HGA3 0.09 !
> ATOM H63 HGA3 0.09 !
> GROUP !
> ATOM CM CG331 -0.27
> ATOM HM1 HGA3 0.09
> ATOM HM2 HGA3 0.09
> ATOM HM3 HGA3 0.09
>
> BOND CM HM1 CM HM2 CM HM3
> BOND N9 C8 N9 C4 C8 N7 C8 H8
> BOND N7 C5 C5 C6 C5 C4 C6 N6
> BOND C6 N1 N6 CM6 N6 H6 N1 C2
> BOND C2 N3 C2 H2 N3 C4 CM6 H61
> BOND CM6 H62 CM6 H63
> BOND CM N9
> IMPR C6 C5 N1 N6 ! N6 C6 CM6 H6
> DONO H6 N6
> ACCE N3
> ACCE N7
> ACCE N1
> IC C8 C4 *N9 CM 1.3830 106.03 179.68 126.91 1.4740
> IC C4 N9 C8 N7 1.3497 107.34 0.02 112.90 1.3249
> IC N7 N9 *C8 H8 1.3249 112.90 -179.86 122.61 1.0947
> IC N9 C8 N7 C5 1.3736 112.90 0.40 103.62 1.3989
> IC C4 N7 *C5 C6 1.3954 110.25 -178.17 133.05 1.4051
> IC N7 C5 C6 N1 1.3989 133.05 177.05 117.60 1.3793
> IC N1 C5 *C6 N6 1.3793 117.60 179.99 122.23 1.3815
> IC C5 C6 N6 CM6 1.4051 122.23 167.61 118.33 1.4832
> IC CM6 C6 *N6 H6 1.4832 118.33 180.00 112.50 1.0190
> IC C5 C6 N1 C2 1.4051 117.60 1.54 120.04 1.3647
> IC C6 N1 C2 N3 1.3793 120.04 -1.14 125.07 1.3600
> IC N3 N1 *C2 H2 1.3600 125.07 -179.44 117.43 1.0938
> IC C6 N6 CM6 H61 1.3815 118.33 174.59 110.48 1.1123
> IC H61 N6 *CM6 H62 1.1123 110.48 120.34 111.22 1.1126
> IC H61 N6 *CM6 H63 1.1123 110.48 -119.06 111.26 1.1131
>
> On Mon, Dec 25, 2017 at 10:17 AM, Lennart Nilsson <Lennart.Nilsson_at_ki.se <mailto:Lennart.Nilsson_at_ki.se>> wrote:
> This residue is defined as a complete residue (RESI 6MA) and not as a patch; in this file patches are mainly used for alternative tautomers or protonation states. Just check the supplementary information of the JCC paper, or why not simply have look at the file you are trying to use – “toppar_all36_na_rna_modified.str”  is a human-readable text file, and it is quite clear that 6MA is a RESIdue and not a patch (PRES): <>
> toppar_all36_na_rna_modified.str:RESI 6MA -1.00 ! N6-methyladenosine
>
> Good luck!
> Lennart Nilsson
>
> From: owner-namd-l_at_ks.uiuc.edu <mailto:owner-namd-l_at_ks.uiuc.edu> [mailto:owner-namd-l_at_ks.uiuc.edu <mailto:owner-namd-l_at_ks.uiuc.edu>] On Behalf Of The Cromicus Productions
> Sent: den 24 december 2017 10:49
> To: NAMD list <namd-l_at_ks.uiuc.edu <mailto:namd-l_at_ks.uiuc.edu>>
> Subject: namd-l: [NAMD] Question about patches
>
> Hi everyone,
>
> I'm working with DNA and trying to convert an adenine group into an
> N6-methyladenosine. I found the force field from
>
> Xu, Y., K. Vanommeslaeghe, A. Aleksandrov, A. D. MacKerell and L. Nilsson (2016). "Additive CHARMM force field for naturally occurring modified ribonucleotides." Journal of Computational Chemistry 37: 896-912
>
> on one of the latest modifications of CHARMM36 (toppar_all36_na_rna_modified.str) but I'm having problems because there is no patch defined to do this modification and I don't know how to define a patch. Could anyone lead me towards how to perform this task?
> I have succeeded before with other modifications such as methylating C bases, but I always had a patch defined in the force field to do this.
>
> Thank you very much and merry Christmas,
>
> Sebastian
>
>
>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:46 CST