[gmx-users] make_ndx error

[Justin Lemkul]

On 1/26/13 1:57 PM, Albert wrote:
> On 01/26/2013 07:51 PM, Justin Lemkul wrote:
>>
>> Can you please post the following:
>>
>> 1. The groups printed in the make_ndx prompt
>> 2. The output of gmxcheck on an index file created from your coordinate file
>> (created simply by typing 'q' at the prompt, i.e. not creating any special
>> groups)
>>
>> -Justin
>
>
> make_ndx -f sys.pdb
>
>   0 System              : 59870 atoms
>    1 Protein             :  4746 atoms
>    2 Protein-H           :  2329 atoms
>    3 C-alpha             :   292 atoms
>    4 Backbone            :   877 atoms
>    5 MainChain           :  1170 atoms
>    6 MainChain+Cb        :  1453 atoms
>    7 MainChain+H         :  1455 atoms
>    8 SideChain           :  3291 atoms
>    9 SideChain-H         :  1159 atoms
>   10 Prot-Masses         :  4746 atoms
>   11 non-Protein         : 55124 atoms
>   12 Other               : 18766 atoms
>   13 NMA                 :     6 atoms
>   14 POPC                : 18760 atoms
>   15 CL                  :    39 atoms
>   16 NA                  :    34 atoms
>   17 Ion                 :    73 atoms
>   18 NMA                 :     6 atoms
>   19 POPC                : 18760 atoms
>   20 CL                  :    39 atoms
>   21 NA                  :    34 atoms
>   22 Water               : 36285 atoms
>   23 SOL                 : 36285 atoms
>   24 non-Water           : 23585 atoms
>   25 Water_and_ions      : 36358 atoms
>
>
>
> gmxcheck_mpi -f md.xtc -n index.ndx
>

Note that -f is not necessary to produce the needed information, just -n.

> Item        #frames Timestep (ps)
> Step           126    0.1
> Time           126    0.1
> Lambda           0
> Coords         126    0.1
> Velocities       0
> Forces           0
> Box            126    0.1
> Contents of index file index.ndx
> --------------------------------------------------
> Nr.   Group               #Entries   First    Last
>     0  System                 59870       1   59870
>     1  Protein                 4746       1    4752
>     2  Protein-H               2329       1    2332

Something is screwy here - note that gmxcheck finds that the index file has all 
of the protein heavy atoms first, then all of its hydrogens.  Is that correct? 
It's certainly not normal for any structure processed by pdb2gmx.  Besides the 
point, I suppose.

>     3  C-alpha                  292       8    2323
>     4  Backbone                 877       2    2324
>     5  MainChain               1170       2    2325
>     6  MainChain+Cb            1453       2    2326
>     7  MainChain+H             1455       2    4751
>     8  SideChain               3291       1    4752
>     9  SideChain-H             1159       1    2332
>    10  Prot-Masses             4746       1    4752
>    11  non-Protein            55124    4745   59870
>    12  Other                  18766    4745   23512
>    13  NMA                        6    4745    4750
>    14  POPC                   18760    4753   23512
>    15  CL                        39   23513   23585
>    16  NA                        34   23518   23551
>    17  Ion                       73   23513   23585
>    18  NMA                        6    4745    4750
>    19  POPC                   18760    4753   23512
>    20  CL                        39   23513   23585
>    21  NA                        34   23518   23551
>    22  Water                  36285   23586   59870
>    23  SOL                    36285   23586   59870
>    24  non-Water              23585       1   23585
>    25  Water_and_ions         36358   23586   23585

Here's the issue.  You've got a Water_and_ions group that has over 36k atoms, 
but it starts by counting backwards!  23586 to 23585 is a size of -1.  When you 
inspect the index file in a text editor, what's in this group?

-Justin

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

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

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