import/export for secondary structures

[westram]

• translate ARBs HELIX+HELIX_NR from/to common format(s)
• See
  • results of  http://mfold.rna.albany.edu/?q=mfold/RNA-Folding-Form
  •  http://en.wikipedia.org/wiki/List_of_RNA_structure_prediction_software

for some ss-format-examples

Problems:

• ARB helix numbering will be lost when doing export+import
• has to be alignment-independent to import correctly after inserting/deleting gaps
  • some formats use one reference sequence (for prediction software this is the sequence on which the prediction has been calculated) - fine for importing that structure
  • when exporting, ARB would need a sequence which is covering all helix-columns (it's quite unlikely that such sequence exists)

Possible solutions:

• cook own reimportable format
  • save helix-names
  • save a set of sequences, such that for each helix column at least one of them has data
    • may still be impossible (if no seq has data at one helix position)
    • can only be imported, if all sequences of the set exists in the target DB

ideas anybody?