Netiquette · Download · News · Gallery · Homepage · DSSR Manual · G-quadruplexes · DSSR-Jmol · DSSR-PyMOL · DSSR Licensing · Video Overview· RNA Covers

Author Topic: Issue with lattice contacts  (Read 12009 times)

Offline Luigi

  • with-posts
  • *
  • Posts: 7
    • View Profile
Issue with lattice contacts
« on: July 27, 2015, 01:56:54 pm »
Hello Xiang-Jun and congratulations for the NAR paper!
I came into a problem using DSSR on several small structures, I will show you 3HTS as example. The input is the following:

x3dna-dssr -i=/home//3HTS.py.pdb --segid --idstr=long --u-turn --more --non-pair --po4 --nested -o=3HTS.py.dssr.out

Here the fragment of the output file which causes me troubles:

****************************************************************************
List of 3 atom-base capping interactions
    dv: vertical distance of the atom above the nucleotide base
    -----------------------------------------------------------
     type       atom                 nt             dv
   1 sugar      O4'@.B000.A.DG.1.    ..A.DC.12.     2.84
   2 sugar      O4'@..A.DG.1.        .A00Z.A.DC.12. 2.84
   3 sugar      O4'@.B000.A.DG.1.    .B00A.A.DC.12. 2.84

****************************************************************************
I attached the output file "dssr-a2bases" to show you better the issue. In fact it seems that in the capping contact between the DG 1A and DC 12A there are some troubles related to the symmetry codes. If I open the file with PyMol generating symmetry mates, I find that the residue DG 1A with symmetry A00A is stacked over DC 12A, while in the dssr output the symmetric B000 is reported. I know it's a quite tricky matter and occurs only in small structures, but I still wanted to show it to you.

Thank you in advance for your answer

Luigi



Offline xiangjun

  • Administrator
  • with-posts
  • *****
  • Posts: 1640
    • View Profile
    • 3DNA homepage
Re: Issue with lattice contacts
« Reply #1 on: July 27, 2015, 02:02:50 pm »
Hi Luigi,

Thanks for reporting the issue. Please attach the "3HTS.py.pdb" file so I (or others who are interested) can reproduce your reported problem.

Note that DSSR is not "smart" by design -- it just checks the input structure as is (actaully the first model terminated by the END or ENDMDL record, by default), without taking into consideration of crystallographic symmetry. It is up to the user to prepare the input struture with all required atomic coordinates for DSSR to work on.

Best regards,

Xiang-Jun
« Last Edit: July 27, 2015, 03:24:59 pm by xiangjun »

Offline Luigi

  • with-posts
  • *
  • Posts: 7
    • View Profile
Re: Issue with lattice contacts
« Reply #2 on: July 28, 2015, 05:26:12 am »
Hi Xiang-Jun,

thanks for the answer. Here's the "3HTS.py.pdb" file used as DSSR input.
Measuring the distance from the B000.A.DG.1. and ..A.DC.12. we get around 40 Å and not 2.84 Å as reported in the output in this line

1 sugar      O4'@.B000.A.DG.1.    ..A.DC.12.     2.84

Best regards,

Luigi

Offline xiangjun

  • Administrator
  • with-posts
  • *****
  • Posts: 1640
    • View Profile
    • 3DNA homepage
Re: Issue with lattice contacts
« Reply #3 on: July 28, 2015, 02:02:12 pm »
Hi Luigi,

Thanks for providing the "3HTS.py.pdb" file which helped me to trace where the problem is. As shown below, the two neighboring A.DG1 nts are differentiated only by segids "B000" and "A00A", which were not taken into account by the DSSR version you used. So both residues are taken as ONE unit, even though they are far off as shown in your initial attached image.
ATOM      1  O5'  DG A   1      39.110  13.077   4.497  1.00 33.08      B000 O  
ATOM      2  C5'  DG A   1      38.049  13.488   3.641  1.00 31.25      B000 C 
ATOM      3  C4'  DG A   1      37.215  12.293   3.246  1.00 29.19      B000 C 
......
ATOM     18  N3   DG A   1      33.391  11.463  -0.413  1.00 20.16      B000 N 
ATOM     19  C4   DG A   1      33.803  12.711  -0.134  1.00 22.14      B000 C 
ATOM      1  O5'  DG A   1       1.555  13.077  30.156  1.00 33.08      A00A O 
ATOM      2  C5'  DG A   1       2.616  13.488  31.012  1.00 31.25      A00A C 
ATOM      3  C4'  DG A   1       3.450  12.293  31.407  1.00 29.19      A00A C 
......
ATOM     18  N3   DG A   1       7.274  11.463  35.066  1.00 20.16      A00A N 
ATOM     19  C4   DG A   1       6.862  12.711  34.787  1.00 22.14      A00A C 

If you paid attention to DSSR diagnostic message, you may have noticed the following:

Code: [Select]
more than one < O5'> atoms in A.DG1
*****the first match is used*****
more than one < C5'> atoms in A.DG1
*****the first match is used*****
        ......

This bug only shows up in very special cases as yours -- with segid info, and two otherwise identical residues next to each other. I've fixed the issue and updated DSSR binaries in the download page (currently still called v1.2.9-2015jul25). Now the corresponding section is as below:
List of 3 atom-base capping interactions
    dv: vertical distance of the atom above the nucleotide base
    -----------------------------------------------------------
     type       atom                 nt             dv
   1 sugar      O4'@.A00A.A.DG.1.    ..A.DC.12.     2.84
   2 sugar      O4'@..A.DG.1.        .A00Z.A.DC.12. 2.84
   3 sugar      O4'@.B000.A.DG.1.    .B00A.A.DC.12. 2.84

Please have a try and report back how it goes.

Best regards,

Xiang-Jun
« Last Edit: July 29, 2015, 11:06:32 am by xiangjun »

Offline Luigi

  • with-posts
  • *
  • Posts: 7
    • View Profile
Re: Issue with lattice contacts
« Reply #4 on: August 17, 2015, 01:03:34 pm »
Hi Xiang-Jun,

I apologize for the delay in the answer. I tried the new version and it works perfectly!

Thank you,

Luigi

 

Created and maintained by Dr. Xiang-Jun Lu [律祥俊] (xiangjun@x3dna.org)
The Bussemaker Laboratory at the Department of Biological Sciences, Columbia University.