[gmx-users] Warning: atom names in CG topology file and CG pdb file do not match

[Justin A. Lemkul]

Faezeh Kargar wrote:
> Dear All
> I searched my problem in users mailing list but I could not find the 
> answer. please help me with this.
> I am learning the CG Martini force filed using the tutorial "protein in 
> water" for protein ubiquitin. so I did the things like below:
> 
> 1. awk -f atom2cg_v2.1.awk 1UBQ.pdb > 1UBQ-cg.pdb (it makes a CG .pdb file)
> 
> 2. wget ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt
> 
> 3. grep -A 1 1ubq pdb_seqres.txt > 1ubq.seq (these 2 and 3 commands make 
> a .seq file)
> 
> 4. ./seq2itp.pl 1ubq.seq 1ubq.ssd > 1ubq.itp (by the use of .seq and 
> .ssd file and seq2itp.pl script I can make a .itp file. I should mention 
> that I used .ssd file prepared by matini group in this tutorial.)
> 
> 5. I make a 1ubq.top file like below
> 
> #include "martini_v2.1.itp"
> #include "1ubq.itp"
> 
> [ system ]
> ; name
> protein
> 
> [ molecules ]
> ; name  number
> protein 1
> 
> 6. grompp -f 1ubq.mdp -c 1UBQ-cg.pdb -p 1ubq.top
> 
> but warnings come like this
> 
> Warning: atom name 1 in 1ubq.top and 1UBQ-cg.pdb does not match (BBc - BB)
> Warning: atom name 2 in 1ubq.top and 1UBQ-cg.pdb does not match (S1c - S1)
> Warning: atom name 3 in 1ubq.top and 1UBQ-cg.pdb does not match (BBe - BB)
> ............
> 
> ............
> 
> all of my atom names mismatch. I checked the 1ubq.itp. some lines are in 
> below
> 
> ;;; MARTINI 2.1 coarse-grained topology
> ;;; Generated by seq2itp.pl version 1.1.4
> 
> ;;; SEQ: MQIFVKTLTG KTITLEVEPS DTIENVKAKI QDKEGIPPDQ QRLIFAGKQL  50
> ;;; SEQ: EDGRTLSDYN IQKESTLHLV LRLRGG
> ;;; Total Number of Amino Acid Residues: 76
> 
> 
> [moleculetype]
> ;molname    exclusions
> Protein    1
> 
> [atoms]
> 1    Qd    1    MET    BBc    1    1.000    ;    COI
> 2    C5    1    MET    S1c    2    0.000    ;    COI
> 3    Nda    2    GLN    BBe    3    0.000    ;    EXT
> 
> ........
> 
> ........
> 
> [ bonds ]
> ;backbone-backbone bonds
> 1    3    1    0.350    400    ;    COI-EXT
> 3    5    1    0.350    1250    ;    EXT-EXT
> 5    7    1    0.350    1250    ;    EXT-EXT
> 
> ........
> 
> ........
> 
> [ constraints ]
> ;sc-sc constraints (Ring Structures)
> 8    9    1    0.270    ;    PHE4
> 8    10    1    0.270    ;    PHE4
> 9    10    1    0.270    ;    PHE4
> 
> .........
> 
> .........
> 
> [angles]
> ;backbone-backbone-backbone angles
> 1    3    5    2    127.00    25    ;    COI-EXT-EXT
> 3    5    7    2    134.00    25    ;    EXT-EXT-EXT
> 5    7    11    2    134.00    25    ;    EXT-EXT-EXT
> 
> .............
> 
> .............
> 
> [dihedrals]
> ;improper dihedral angles
> 7    9    10    8    2    0.00    50    ;    PHE4
> 94    96    97    95    2    0.00    50    ;    PHE45
> 
> ..........
> 
> ..........
> 
> I found that for example the word "c" in atom name "BBc" comes from 
> "COI" means coil. so I added manually such these words to my 1UBQ-cg.pdb 
> file acording to 1ubq.itp file. but then again the same warnings. Please 
> help me with these too many warnings.
> 

It's been a while since I did the MARTINI tutorial, but as I recall, the grompp 
warnings about names are to be expected and are actually necessary to get the 
right atom names.  This is one of the few times when running grompp with 
-maxwarn is acceptable, and in fact, beneficial.

> 7. I used -maxwarn, but when I used the command "mdrun" the message like 
> this appears:
> 

There are numerous reasons why LINCS warnings appear.  Please refer to the 
following:

http://www.gromacs.org/Documentation/Terminology/Blowing_Up

Without seeing your .mdp file, or the results of any prior energy minimization, 
etc it is impossible to say what the problem is definitively.  I see you're 
trying to use a 40-fs timestep; I have never been able to run a stable MARTINI 
simulation with a timestep greater than 20 fs, but I have no idea why.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================