brackets in DNA-RNA pairing

[ICdB]

Hi,

I ran dssr locally on PDB 4OO8, which contains 2 symmetrical copies of a RNA-DNA duplex, made of chains E-F and B-C. The only difference in the 2 duplexes is 3 additional bases in 5' of the DNA chain.
However, the secondary structure given for each duplex is different:

for DNA - RNA:
chain_C['sstr'], chain_B['sstr'] =  '[[[[[[[[[[[[[[[[[[[[',  ']]]]]]]]]]]]]]]]]]]]'
chain_F['sstr'], chain_E['sstr'] = '((((((((((((((((((((' , '))))))))))))))))))))'

When running dssr on http://jmol.x3dna.org/, the secondary structure is identical for the 2 duplexes (with round brackets).
So what is the meaning of the squared brackets in the local version?

Thanks in advance,
Isaure

[xiangjun]

Hi Isaure,

Thanks for using DSSR and for posting your message on the 3DNA Forum.

Running DSSR on PDB entry 4oo8, I got the following:

Code: [Select]

****************************************************************************
Secondary structures in dot-bracket notation (dbn) as a whole and per chain
>4oo8 nts=236 [whole]
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&GCCAAGCGCACCTAATTTCC&GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&CAGCCAAGCGCACCTAATTTCC
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&))))))))))))))))))))&((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&..))))))))))))))))))))
>4oo8-B #1 nts=97 0.32(2.95) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4oo8-C #2 nts=20 2.66(1.04) [chain] DNA
GCCAAGCGCACCTAATTTCC
))))))))))))))))))))
>4oo8-E #3 nts=97 0.39(2.98) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4oo8-F #4 nts=22 2.61(1.34) [chain] DNA
CAGCCAAGCGCACCTAATTTCC
..))))))))))))))))))))
Could you please provide details on how you ran DSSR command-line on it to get the output you reported?

Best regards,

Xiang-Jun

[ICdB]

Hi Xiang-Jun,

Thanks for your prompt answer.
Sorry for a mistake in my question. I ran the local version on a modified pdb (attached) with only the 2 duplexes (wo. the protein) and wo. the 3 additional nucleotides 5' from chain F. When running dssr online or locally on that pdb, I get the same results : squared / round brackets for the B-C / E-F duplex respectively. 
So now my question is about the difference of treatment of those 2 duplexes.

Thanks again in advance,
Isaure

[xiangjun]

Hi Isaure,

Thanks for your followup. I've run DSSR on your attached 4OO8-noprot.pdb, and got the following:

Code: [Select]

Secondary structures in dot-bracket notation (dbn) as a whole and per chain
>4OO8-noprot nts=236 [whole]
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&GCCAAGCGCACCTAATTTCC&GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&CAGCCAAGCGCACCTAATTTCC
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&))))))))))))))))))))&((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&..))))))))))))))))))))
>4OO8-noprot-B #1 nts=97 0.32(2.95) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-noprot-C #2 nts=20 2.66(1.04) [chain] DNA
GCCAAGCGCACCTAATTTCC
))))))))))))))))))))
>4OO8-noprot-E #3 nts=97 0.39(2.98) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-noprot-F #4 nts=22 2.61(1.34) [chain] DNA
CAGCCAAGCGCACCTAATTTCC
..))))))))))))))))))))
So I still did not see the problem you reported. What version of DSSR are you using?

Xiang-Jun

[ICdB]

Hi Xiang-Jun,

I use version v1.7.2-2017nov20

I did some more testing, and found that identical PDB files but for the order of the chains get different results. On the two pdb attached, I get this:

Secondary structures in dot-bracket notation (dbn) as a whole and per chain
>4OO8-order1 nts=234 [whole]
GCCAAGCGCACCTAATTTCC&GCCAAGCGCACCTAATTTCC&GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((&[[[[[[[[[[[[[[[[[[[[&))))))))))))))))))))((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&]]]]]]]]]]]]]]]]]]]]((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-order1-C #1 nts=20 2.66(1.04) [chain] DNA
GCCAAGCGCACCTAATTTCC
((((((((((((((((((((
>4OO8-order1-F #2 nts=20 2.67(1.05) [chain] DNA
GCCAAGCGCACCTAATTTCC
[[[[[[[[[[[[[[[[[[[[
>4OO8-order1-B #3 nts=97 0.32(2.95) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
))))))))))))))))))))((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-order1-E #4 nts=97 0.39(2.98) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
]]]]]]]]]]]]]]]]]]]]((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).

Secondary structures in dot-bracket notation (dbn) as a whole and per chain
>4OO8-order2 nts=234 [whole]
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&GCCAAGCGCACCTAATTTCC&GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU&GCCAAGCGCACCTAATTTCC
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&))))))))))))))))))))&((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).&))))))))))))))))))))
>4OO8-order2-B #1 nts=97 0.32(2.95) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-order2-C #2 nts=20 2.66(1.04) [chain] DNA
GCCAAGCGCACCTAATTTCC
))))))))))))))))))))
>4OO8-order2-E #3 nts=97 0.39(2.98) [chain] RNA
GGAAAUUAGGUGCGCUUGGCGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCCGUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGCU
((((((((((((((((((((((((((..((((....))))....))))))..(((..).)).......((((....)))).((((((...)))))).
>4OO8-order2-F #4 nts=20 2.67(1.05) [chain] DNA
GCCAAGCGCACCTAATTTCC
))))))))))))))))))))

[xiangjun]

Hi Isaure,

Thanks for your diagnosis of the reported issue.

I did some more testing, and found that identical PDB files but for the order of the chains get different results.

This is the expected behavior, and what I thought when I first read your post. When deriving the DBN, DSSR assumes the given structure, as a whole, is as if in a continuous chain. Whenever there is a break (between chains or fragments within a chain), an '&' symbol is introduced. When a structure contains more than one chain, they can be ordered differently. DSSR follows the input order strictly, by design. There are tools available to put the chains in 'canonical' order to remove such false 'pseudoknots'.

Best regards,

Xiang-Jun

[xiangjun]

Also, pay attention to DSSR output, as shown below for "4OO8-order1.pdb"

****************************************************************************
Special notes:
   o cross-paired segments in separate chains, be *careful* with .dbn

****************************************************************************
This structure contains *1-order pseudoknot
   o You may want to run DSSR again with the '--nested' option which removes
     pseudoknots to get a fully nested secondary structure representation.
   o The DSSR-derived dbn may be problematic (see notes above).

[ICdB]

Hi Xiang-Jun,

ok, thanks a lot for your help and your detailed answers. And sorry for missing the warning in output.

Best regards,
Isaure

[xiangjun]

Hi Isaure,

You're welcome. I'm glad that you raised up this issue. As a side note, you could display the secondary structure using VARNA using DSSR-derived dssr-2ndstrs.ct or dssr-2ndstrs.bpseq file (in addition to dssr-2ndstrs.dbn), and you would see the same 'issue'.

In fact, this DBN issue is only one of that many subtle points that are not 'documented'. They are too technical in nature, and too detailed for the majorities of DSSR users. So, for practical considerations, I've decided to address them on a case-by-case base (only) when asked.

Best regards,

Xiang-Jun