Calculating the helical axis of DNA

[rcrehuet]

Dear all,
I would like to extract the DNA helical axis from a pdb, which contains a protein and DNA.
I first run

Code: [Select]

find_pairs mypdb.pdb sample.inpThen, I guess I should run analyze (am I correct?):

Code: [Select]

analyze sample.inpEven though I get the following error message some files are generated.

Code: [Select]

......Processing structure #1: <sample.inp>......
[i] missing ' P  ' atom : residue name 'DA5', chain D, number [   1 ]
[i] missing ' OP1' atom : residue name 'DA5', chain D, number [   1 ]
[i] missing ' OP2' atom : residue name 'DA5', chain D, number [   1 ]
[i] missing ' P  ' atom : residue name 'DC5', chain E, number [   1 ]
[i] missing ' OP1' atom : residue name 'DC5', chain E, number [   1 ]
[i] missing ' OP2' atom : residue name 'DC5', chain E, number [   1 ]
[i] missing ' P  ' atom : residue name 'DA5', chain D, number [   1 ]
[i] missing ' P  ' atom : residue name 'DC5', chain E, number [   1 ]
My question is very simple. Where can I find the global linear helical axis in terms of its x, y, z coordinates?
At the end of the out file I see "local helical axis vector" where there is a vector for each pair. Should I average these vectors? I thought the axis was calculated by a fit to all base pairs, and so that there would be a single axis vector, not one for each base pair. Did I get something wrong?
Thanks in advance!
Ramon

[xiangjun]

Hi Ramon,

Thanks for using 3DNA! The way you run the program,

Code: [Select]

find_pairs mypdb.pdb sample.inp
analyze sample.inp
is correct. In 3DNA v2.3, you could also combine the two steps into one, as below:

Code: [Select]

find_pairs mypdb.pdb | analyze
In all cases, the main output file will be in file 'mypdb.out'.

Even though I get the following error message some files are generated.

Code: [Select]

......Processing structure #1: <sample.inp>......
[i] missing ' P  ' atom : residue name 'DA5', chain D, number [   1 ]

As the symbol 'i' implies, this is not an error message, but an informational diagnostic message. It has clear meaning, i.e. DA5 (residue #1) does not have a P atom in it.

My question is very simple. Where can I find the global linear helical axis in terms of its x, y, z coordinates?

Using 355d as an example, with the following command:

Code: [Select]

find_pair 355d.pdb | analyze
You will find the following section in 355d.out:

Code: [Select]

****************************************************************************
Global linear helical axis defined by equivalent C1' and RN9/YN1 atom pairs
Deviation from regular linear helix: 3.30(0.52)
Helix:    -0.1269   -0.2753   -0.9530
HETATM 9998  XS    X X 999      17.536  25.713  25.665
HETATM 9999  XE    X X 999      12.911  15.677  -9.080
Average and standard deviation of helix radius:
      P: 9.42(0.82), O4': 6.37(0.85),  C1': 5.85(0.86)
Note that if your input structure is too curved, this section will not be output by default. If you *still* want to have such an output (which may not be that meaningful), you need to change the 'std_curved' field in file $X3DNA/config/misc_3dna.par:

Code: [Select]

# Section 4: is this double helix curved?
#   criterion to decide if a helix is strongly curved.
#   if relatively straight, 'analyze' will output a set of
#   global parameters, the normalized axis vector, and the
#   two end points the helical axis passes through
<std_curved>0.6</std_curved>

E.g., change 0.6 to 1.0.

Search the Forum for related topics. You may also want to repeat Recipe#4 of the 2008 3DNA Nature Protocols paper.

Alternatively, you may give DSSR a try where the helical axis info is alway output with the --more option, as in:

Code: [Select]

x3dna-dssr -i=355d.pdb --more
The above command will output:

Code: [Select]

  helix#1[1] bps=12
      strand-1 5'-CGCGAATTCGCG-3'
       bp-type    ||||||||||||
      strand-2 3'-GCGCTTAAGCGC-5'
      helix-form  BBBBBBBBBBB
    helical-rise:   3.30(0.52)
    helical-radius: 9.42(0.82)
    helical-axis:   -0.127    -0.275    -0.953
       point-one:   17.536    25.713    25.665
       point-two:   12.911    15.677    -9.080

If the std of helical rise (the number in parenthesis following helical-rise, 0.52 in the 355d example), a * will be added to draw user's attention.

HTH,

Xiang-Jun

[rcrehuet]

Hi Xiang-Jun,
Thanks a lot for your explanation!!
Best,
Ramon

[rcrehuet]

Hi again,
I have a further question to this thread...
If I'm analysing a multi-model PDB, I followed this steps:

Code: [Select]

find_pair out.pdb out.inp
x3dna_ensemble extract -p hx_vec -o helix.out
This gives me the local helical axis. Is there a keyword (or an alternative procedure) to get the global linear helical axis for all the PDB models?
Thanks again!
Ramon

[xiangjun]

Hi Ramon,

The proper way of using the x3dna_ensemble Ruby script is:

Code: [Select]

       x3dna_ensemble analyze -h
       x3dna_ensemble extract -h
Unfortunately, x3dna_ensemble currently does *not* extract the global linear helical axes for all the PDB models. If you are interested, you may have a look of $X3DNA/lib/parser.rb and try to add the feature yourself.

Xiang-Jun