interaction code and other issues

[auffinger]

Hi Xiang-Jun,

Really great work, thanks.

Just a few remarks:
Why didn't you use the same nomenclature as in dssr ?
I mean using dots to separate the nucleotide name from its numbering.
For example in 4LT5, I see

****************************************************************************
List of 6 pair-wise phosphate-group/amino-acid H-bonding interactions
      id  nt-aa  nt           aa          H-bonds
   1 4LT5 G-lys B.DG9        A.LYS318     1:OP1-NZ[3.42]

B.DG9 could be B.DG.9
A.LYS318 could be A.LYS.318

Have you taken into account the segid code as you did for dssr ?

for the same structure, we defined a 3.5 cut-off
Command: x3dna-snap -i=/media/HD/DATA/pdb_files_temp/4LT5/4LT5.py.pdb -o=4LT5.py.snap.3.5.out --cutoff=3.5

yet we found
  2 4LT5 T-lys B.DT11       A.LYS303     1:OP1-NZ[3.71]  <---- greater than 3.5
   1 4LT5 T-lys B.DT10       A.LYS303     1:O3'-NZ[3.65]  <---- greater than 3.5

Thanks for looking at this,

Pascal

[xiangjun]

Hi Pascal,

Thanks for being the first to post SNAP-related questions/comments on the Forum!

Why didn't you use the same nomenclature as in dssr ?
I mean using dots to separate the nucleotide name from its numbering.
For example in 4LT5, I see

****************************************************************************
List of 6 pair-wise phosphate-group/amino-acid H-bonding interactions
      id  nt-aa  nt           aa          H-bonds
   1 4LT5 G-lys B.DG9        A.LYS318     1:OP1-NZ[3.42]

B.DG9 could be B.DG.9
A.LYS318 could be A.LYS.318

Have you taken into account the segid code as you did for dssr ?

SNAP does use the same nomenclature as in DSSR. Specify the --long-idstr to get dot-separated id string; with --segid, you also get the segid code in the identifier. For example, you will have the following result for 4lt5 with --long-idstr:

  1 4lt5 G-lys .B.DG.9.     .A.LYS.318.  1:OP1-NZ[3.42]

for the same structure, we defined a 3.5 cut-off
Command: x3dna-snap -i=/media/HD/DATA/pdb_files_temp/4LT5/4LT5.py.pdb -o=4LT5.py.snap.3.5.out --cutoff=3.5

yet we found
  2 4LT5 T-lys B.DT11       A.LYS303     1:OP1-NZ[3.71]  <---- greater than 3.5
   1 4LT5 T-lys B.DT10       A.LYS303     1:O3'-NZ[3.65]  <---- greater than 3.5

The discrapnacy you noticed is due to the two different distance cutoffs: the --cutoff=3.5 with SNAP is the contact distance threshold with a default of 4.5 angstrom, while the default H-bond distance cutoff is 4.0 as in DSSR (and 3DNA).

HTH,

Xiang-Jun

[auffinger]

Thanks Xiang-Jun,

For h_bonds in snap, can I use --hbond_d2=3.5 as in dssr ?

Best,

Pascal

[xiangjun]

Yes. Have a try and report back if you found otherwise -:)

Xiang-Jun

[auffinger]

Hi Xiang-Jun,

I use the  following options:
x3dna-snap --long-idstr --segid --cutoff=3.5 --hbond_d2=3.5 -i=xxx -o=yyy

It works fine for a lot of files including 4LT5
Yet could you check 1JJ2 or 1FFK that are ribosomes ?

here is my output file

****************************************************************************
    SNAP: a program for the characterization of three-dimensional
             Structures of Nucleic Acid-Protein complexes
        beta-r02-2014may31, Xiang-Jun Lu (xiangjun@x3dna.org)

  This program is being actively maintained and developed. As always,
  I greatly appreciate your feedback! Please report all SNAP-related
  issues on the 3DNA Forum (forum.x3dna.org). I strive to respond
  *promptly* to *any questions* posted there.

****************************************************************************
Note: Each nucleotide/amino-acid is identified by model:chainId.name#, where
      the 'model:' portion is omitted if no model number is available (as is
      often the case for x-ray crystal structures in the PDB). So a common
      example would be B.DA1689, meaning adenosine #1689 on chain B.

Command: x3dna-snap -i=/media/HD/DATA/pdb_files_temp/1FFK/1FFK.py.pdb -o=1FFK.py.snap.3.5.out --cutoff=3.5
Date and time: Mon Jun  2 15:32:22 2014
File name: 1FFK.py.pdb
    no. of peptide chains: 27 [1=14,A=83,B=96,C=59,D=22,E=8,F=22,G=23,H=20,I=57,J=41,K=40,L=23,M=46,N=32,O=40,P=7,Q=25,R=7,S=8,T=25,U=15,V=37,W=14,X=27,Y=11
,Z=38]
    no. of DNA chains: 0 []
    no. of amino acids:    840
    no. of nucleotides:    2838
    no. of atoms:          61687
    no. of waters:         11
    no. of metals:         3 [Mg=2,K=1]
sws2@jabba:/media/HD/DATA/pdb_files_temp/1FFK$
*********

Thus, there are no contacts which seems strange.
(I attached the pdb file we used)

Also as you can see, the options we use are not recapitulated in the command line.
is that what you want ?

Best,

Pascal

[xiangjun]

Hi Pascal,

As noted in the opening post:

The program is currently in beta testing release, focusing on DNA-protein interactions.

For the ribosomes 1jj2 or 1ffk, the expected output would be:

tmp [551] x3dna-snap -i=1ffk.pdb -o=xxx

Processing file '1ffk.pdb' [1ffk]
no DNA chains!

Time used: 00:00:00:02

SNAP found that this entry does not contain a DNA chain, so it exits without further processing. I am a bit surprised that SNAP outputs more info for your 1FFK.py.pdb file. Could you attach the file so I can have a try?

Also as you can see, the options we use are not recapitulated in the command line.
is that what you want ?

This is expected behavior right now.

Xiang-Jun

[xiangjun]

Hi Pascal,

Thanks for sending me the file 1FFK.py.pdb. As shown below, my SNAP run on the file has the expected behavior mentioned in my previous reply.

x3dna-snap -i=1FFK.py.pdb -o=xxx --long-idstr --segid --cutoff=3.5 --hbond_d2=3.5

Processing file '1FFK.py.pdb' [1FFK.py]
no DNA chains!

Time used: 00:00:00:01

To emphasize: SNAP currently works for DNA-protein complexes, but not RNA-protein complexes. The topic on DNA-protein interactions is already quite significant. RNA-protein interactions are, of course, very important and quite interesting to work on. Given the constraints of time and resources, however, I am focusing SNAP on DNA-protein interactions.

Best regards,

Xiang-Jun
 

[auffinger]

OK Xiang-Jun,

So when can the RNA-protein version be expected ?

Best,

Pascal

[xiangjun]

Certainly not in 2014.

On top of time and resources restraints in writing code, I must devote more effort to writing papers as well.

Xiang-Jun

[miaozhichao]

Hi Xiangjun,
I hope you are doing well!
I just found you have such good program.
I also would like to analyze protein-NA interactions.
And the same question: When will the protein-RNA part be available.
I think the H-bond part is very helpful for analyzing the specificity, which is what I want to know.

Best
Zhichao.

[xiangjun]

Hi Zhichao,

Thanks for stopping by the Forum.

SNAP already works on protein-RNA complexes by default. Did you notice anything otherwise? Please be specific.

Regarding H-bonds, by defaut, SNAP outputs only the H-bonds at the interface between protein and DNA or RNA. As in DSSR, it should also have a --get-hbond option to output all H-bonds (I believe -- I do not have access to my computer right now to be sure). In any event, DSSR and SNAP can detect all H-bonds within a given structure, and output to a simple text file or to PyMOL or Jmol for easy visualization.

Best regards,

Xiang-Jun

[miaozhichao]

Hi Xiangjun,
 What great news! I did not find this update previously.
 I am recently considering some specificity related interactions. And I think your SNAP could be perfect in analyzing the structures and use as standard.
I will definitely try it in my new project. I will let you know my results.

Thank you very much!

Zhichao.

[xiangjun]

Hi Zhichao,

Glad to hear that SNAP could of some value to your project. Just remember that I stand firmly behind the program, and aim to respond promptly and concretely to any user questions.

With the publication of DSSR, SNAP has become the next major project I will be gradually focusing on. I welcome your feedback in order to make it a more relevant software product to the community.

Best regards,

Xiang-Jun

PS. I read your two recent publications and like them quite a bit!