Single-Stranded Based Zp Parameter

[slaw]

Xiang-Jun,

I read your recent post about this parameter and was intrigued by Richardson's observation.  Thus, I download v2.1 today (32-bit Ubuntu version) in order to calculate this new Zp parameter.  However, when I run 3DNA on the test case that you discussed (1JJ2.pdb) I can't seem to find where the new parameter is listed.  I see the old dinucleotide Zp parameter but they obviously don't correspond to the values that you had shown (1.9 Angstroms and 4.4 Angstroms).  Perhaps binaries have not been updated yet?

Also, if I'm only interested in the single-stranded based Zp value then can I safely assume that I can simply extract the G175 and U176 residues out from the large PDB structure and still be able to calculate the Zp values.  In other words, since it doesn't rely on base pair information then I don't need the rest of the structure.  Is that correct? 

Finally, since your new single-stranded based Zp parameter does not produce values that exactly match those from the Richardson definition (for the 1JJ2 structure: 2.2 Angstroms and 4.6 Angstroms using the Richardson definition vs. your definition - see above), does that mean that I would need to use a slightly different cutoff value instead of 2.9 Angstroms?  From the 1JJ2 example, perhaps using a value of 3.15 Angstroms is better for 3DNA (1.9 A - 2.2 A = -0.3 A, 4.4 A - 4.6 A = -0.2 A)?

Thank you for your time.

[xiangjun]

Hi,

Thanks for your quick catch -- I am planning to write a post talking about these new parameters :-).

Okay, the new option is -torsion (can be abbreviated to -t) for the analyze program. It is meant to be easy to use, efficient, and robust for real-world applications. As always, the point is best illustrated with concrete examples -- do the following and then check output files '1jj2.tor' and '6tna.tor':

Code: [Select]

analyze -t=1jj2.tor 1jj2.pdb
analyze -torsion=6tna.tor 6tna.pdb
Try to delete base atoms, some of backbone atoms, or use a PNA structure etc, the program should behave properly. I'd like to hear your feedback on how the program is working ...

Xiang-Jun

[slaw]

Xiang-Jun,

Thanks for the quick reply.  The 1JJ2 test case works and I can now see the proper output.  I tried to add this to my 3dna.pl Perl script but quickly found out that a command like the following doesn't work:

find_pair PDB.file OUT.file
analyze -t=test.tor -c OUT.file

The test.tor file doesn't write (and possibly doesn't calculate) anything aside from some title information.  I'm guessing that the "-t" and "-c" options cannot be used together.

However, this works

analyze -t=test.tor PDB.file

[slaw]

To be clear, the PDB file that I am using only contains RNA residues 175-177 and no other atoms or residues.

[xiangjun]

Yes, the -torsion option of analyze is intended to be standalone, starting directly from a PDB file, without first running find_pair. The "-c" option does not apply here and is silently ignored.

As I mentioned in several occasions, all nucleic acid backbone torsion angles can be calculated as follows (using 6tna as an example, and the output file would be in '6tna.outs'):

Code: [Select]

find_pair -s 6tna.pdb 6tna.nts
analyze 6tna.nts

# or the above two commands can be combined:
find_pair -s 6tna.pdb stdout | analyze stdin
This functionality has been there since 3DNA v1.5, but it is not widely recognized/used by the community. Once in a while, I notice from literature (mostly on RNA structures) some sort of ad hoc scripts to calculate just those torsion angles.

Now by adding this -torsion option explicitly, by cutting one extra step, and by providing more useful features, I am hoping the new "analyze -torsion" combo would gradually find its way in RNA structure related research.

As always, I am open to suggestions to continuously refine this new feature of 3DNA.

Xiang-Jun

[slaw]

Xiang-Jun,

Thanks for your explanation(s).  Everything looks okay from here based on the tests that I've done.  I've also successfully added this option to my 3dna.pl Perl script.  I'm going to begin analyzing some simulations and hopefully be able to find more things to say.