3DNA Forum

Questions and answers => RNA structures (DSSR) => Topic started by: shuxiang on June 21, 2019, 02:32:10 pm

Title: strange sequence and dot-bracket notation for dssr output
Post by: shuxiang on June 21, 2019, 02:32:10 pm: Hi Xiang-Jun,

We just came across a strange sequence and dot-bracket notation for the dssr output.
open http://dssr.x3dna.org/ , type in 4v91, go the Secondary structures in dot-bracket notation section. You will see something as following:

>4v91 nts=3482 [whole]
U&U&GACCUCA&AA&UCAGGUAGGAGUACCCG&CUGA&AC&UUAAGCAU&AUCAAUAAGCG&G

A lot of "&"s are inserted in sequence and dot-bracket notation. It looks like a bug for the output.

Best,
Shuxiang
Title: Re: strange sequence and dot-bracket notation for dssr output
Post by: xiangjun on June 21, 2019, 03:21:43 pm: Hi Shuxiang,

Quote
A lot of "&"s are inserted in sequence and dot-bracket notation. It looks like a bug for the output.

It is not a bug, but a feature in DSSR. Read "3.18.12 The --dbn-break option" in the DSSR manual.

For this special case, the atomic model is in bad geometry (even though it has a nominal resolution of 3.7 Å). See the attached image for nucleotides 1-3 (GGU) of chain 3. This is sort of like the 3DNA-rebuilt models with an approximate backbone conformation that needs to be optimized (using Phenix as shown in our web 3DNA 2.0 paper (https://doi.org/10.1093/nar/gkz394)).

You could use "--dbn-break=no" to remove the & symbols, as documented in the DSSR manual.

Best regards,

Xiang-Jun
Title: Re: strange sequence and dot-bracket notation for dssr output
Post by: shuxiang on June 21, 2019, 03:31:57 pm: Hi, Xiang-Jun,

Thank you for your reply. I already added this option for the command-line based analysis.

Best :),
Shuxiang