3DNA Forum

Questions and answers => RNA structures (DSSR) => Topic started by: amirtaghavi on July 26, 2019, 07:23:15 pm

Netiquette · Download · News · Gallery · Homepage · DSSR · Web-DSSR · DSSR Manual · G-quadruplexes · DSSR-Jmol · DSSR-PyMOL

Title: --nmr option with --get-hbond & --analyze
Post by: amirtaghavi on July 26, 2019, 07:23:15 pm
Hello Dr. Xiang-Jun,

I have a pdb file which contains 100 frames. I have used the --non-pair option with --nmr to analyze all frames and got the output for each frame but using the same combination of commands for --get-hbond & --analyze dose not work with --nmr.

The command is:
x3dna-dssr -i=comb.pdb --get-hbond --nmr -o=Hbonds_analysis.dat

--get-hbond is overwritten by --nmr and no H-bond analysis is done (the same with --analyze). So the following is the header of stacking analysis:

Command: x3dna-dssr -i=md_comb.pdb --non-pair --nmr -o=stacking_analysis.txt

File name: md_comb.pdb

and the following is for H-bond analysis:     

Command: x3dna-dssr -i=md_comb.pdb --nmr -o=Hbonds_analysis.txt
Date and time: Fri Jul 26 18:50:59 2019
File name: md_comb.pdb

though I have used the --get-hbond option it does not appear in the output provided by DSSR and no h-bond analysis is done.

So I was wondering if you could help me with my problem.

Thanks in forwrad.

Title: Re: --nmr option with --get-hbond & --analyze
Post by: xiangjun on July 26, 2019, 08:42:52 pm
Hi Amir,

The H-bonding info of each frame/model of the ensemble is available via the --json option. Please have a try and report back how it goes.

Title: Re: --nmr option with --get-hbond & --analyze
Post by: amirtaghavi on July 27, 2019, 09:20:30 am
Thanks a lot, the ensemble option works for --json. Although data extraction was easier with standard x3dna out put (like what I get in stacking with the --nmr  --non-piar combination).   

Title: Re: --nmr option with --get-hbond & --analyze
Post by: xiangjun on July 27, 2019, 10:00:53 am

Glad to hear that the --json option with --nmr does the trick for you. I won't argue with you (or any DSSR user) about the best approach to a particular use case. Overall, though, the structured JSON output is the way to go for parsing data -- it is the USB connector between different parts (programs). In my knowledge of DSSR usages, the JSON output is indeed the preferred way to take advantage of what DSSR has to offer. For example, the DSSR-Jmol interface relies on JSON format. As a more up-to-date example, the DSSR R wrapper from "VeriNA3d: an R package for nucleic acids data mining" (https://doi.org/10.1093/bioinformatics/btz553) employes JSON output as well -- see my blogpost "R wrapper to DSSR in VeriNA3d (http://home.x3dna.org/highlights/r-wrapper-to-dssr-in-verina3d)".

The --nmr and --non-pair combination works as you expected because non-pairing interactions are part of DSSR output. The DSSR output was initially in text format only before the JSON option was introduced. The --get-hbond option, on the other hand, is auxiliary and internal to DSSR. After all, there are numerous other ways to detect H-bonds, including the dedicated tools HBPlus and HBExplore, for example. I'm glad to see the community like this 'extra' feature from DSSR, presumably due to its robust performance in real-world applications.

As a side note, the DSSR JSON output actually contains more information about H-bonds than that from the text output. I have no doubt that these features would be put into good use by the community someday.

Best regards,


Created and maintained by Dr. Xiang-Jun Lu [律祥俊], Principal Investigator of the NIH grant R01GM096889
Dr. Lu is currently affiliated with the Bussemaker Laboratory at the Department of Biological Sciences, Columbia University.