Problem using --pair-list options

[angel]

Dear Dr. Xiangjun,

First of all thank you for developing this software, but we are having some troubles using the --pair-list options. We have tried the examples presented on the manual but doesn't work neither (using last version of X3DNA). We have a non-canonical RNA structure with some mismatches U-U that aren't recognised automatically, so we are using the --pair-list to generate and modify it. The generation works well, the problem is at the moment to introduce through:
       x3dna -dssr -i=1msy.pdb  --pair-list -inpfile=dssr-pairs.txt
It says:
       open_file <dssr-pairs.txt> failed: No such file or directory

And deletes the dssr-pairs.txt file. We would be so gratefull if you know why is giving this problem and not working correctly.

The commands used are:
x3dna-dssr -i=1msy.pdb -o=prueba.out --pair-list (to generate the pair-list)
x3dna -dssr -i=1msy.pdb  --pair-list -inpfile=dssr-pairs.txt  (to analyse it)

[xiangjun]

Hi angel,

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

First of all thank you for developing this software, but we are having some troubles using the --pair-list options. We have tried the examples presented on the manual but doesn't work neither (using last version of X3DNA).

As documented in the DSSR manual, there are three variations to the DSSR --pair-list option:

• By simply specifying the --pair-list option, DSSR will output an additional auxiliary file, named dssr-pairs.txt by default.
  x3dna-dssr -i=1msy.pdb -o=1msy.out --pair-list # --> 'dssr-pairs.txt'
• The name of output pairs list can be customized via the --pair-list-outfile=NAME option, as shown below.
  x3dna-dssr -i=1msy.pdb -o=1msy.out --pair-list-outfile=my-pairs.txt
• User-supplied list of pairs can be specified via the --pair-list-inpfile=NAME option. In such cases, DSSR will use these customized bps instead of identifying them ab initio.
  x3dna-dssr -i=1msy.pdb --pair-list-inpfile=my-pairs.txt

There should be no space between --pair-list and -inpfile in the --pair-list-inpfile=NAME option. Otherwise, DSSR would take --pair-list as an option (the first case) and -inpfile as another option. The default dssr-pairs.txt file should not be used as the input file name via --pair-list-inpfile=NAME.

The generation works well, the problem is at the moment to introduce through:
       x3dna -dssr -i=1msy.pdb  --pair-list -inpfile=dssr-pairs.txt
It says:
       open_file <dssr-pairs.txt> failed: No such file or directory

I'm assuming that in the example, x3dna -dssr means x3dna-dssr, and --pair-list -inpfile=dssr-pairs.txt means --pair-list-inpfile=dssr-pairs.txt (i.e., no spaces). The following commands work as expected:

Code: Text

1. x3dna-dssr -i=1msy.pdb  --pair-list-outfile=my-pairs.txt
2. # edit 'my-pairs.txt' as needed
3. x3dna-dssr -i=1msy.pdb  --pair-list-inpfile=my-pairs.txt

The problem is with the default file dssr-pairs.txt used as input. I'll update the user manual shortly. Any suggestions on how to make the point clear, based on your experience?

We have a non-canonical RNA structure with some mismatches U-U that aren't recognised automatically

Could you share with us an example U-U pair that is not automatically recognized by DSSR? A fragment with atomic coordinates of such an U-U pair would be fine. You may also send me such a data file by email. I'm interested in improving the default DSSR options in as many cases as pragmatically applicable.

Best regards,

Xiang-Jun

[xiangjun]

As a follow-up, I've updated the DSSR User Manual with a caveat on the default file name dssr-pairs.txt with --pairs-list option variants. Hopefully, the instruction is now clear enough so that users would not employ the default OUTPUT file name for INPUT to DSSR.

Even before the update, the DSSR manual deliberately used the name my-pairs.txt with --pair-list-outfile and --pair-list-inpfile (as shown in my previous response). However, the message was presumably too subtle to be noticed by first-time users.

Best regards,

Xiang-Jun

[angel]

It finally worked thanks to the change of the input name. Thanks Xiangjun for the fast and precise response, im going to share you via e-mail the pdb file of our structure so you can check the base pair recognision, but it works fine with the --pair-list option.

Best regards,
Ángel.

[xiangjun]

Hi Ángel,

It finally worked thanks to the change of the input name.

Glad to hear that DSSR works as expected with a change of the input file name to the --pairs-list-inpfile option.

Each time DSSR runs, it remove a long list of files (e.g., dssr-pairs.pdb, dssr-multiplets.pdb etc.) that can be automatically created by default. This design ensures a clean state for each DSSR run. As a result, users should name an input file differently than the default to avoid it being removed or overwritten.

Thanks Xiangjun for the fast and precise response, im going to share you via e-mail the pdb file of our structure so you can check the base pair recognision, but it works fine with the --pair-list option.

I've quickly checked the PDB file you emailed me. DSSR is working as expected for your structure: using the U11-U38 'pair' as an example, they are too far apart to be termed a pair in the conventional sense. The shortest O...O or O...N distances is around 6 angstrom, much longer than even the generous 4.0 A cutoff for H-bond used in DSSR. That's why DSSR does not detect them. The DSSR --pair-list options have been designed precisely for a usecase like yours.

Best regards,

Xiang-Jun