1.6.0 SHUFFLE_PADS settings in .gaprc

2005-12-29
2013-04-18
  • So are these two gaprc values deprecated in the new 1.6.0 version of the staden pkg release? I was using these settings to drop columns that were partially padded. And our finishers have gotten used to using this feature.  They all have CONTIG_EDITOR.SHUFFLE_PADS.CONSENSUS_CUTOFF  .75 in their .gaprc files.

    Seems the new version of shuffle pads in 1.6.0 only drops full columns of pads (100%).  Is there some other way to still do what the older version of shuffle pads did?

    # Modifying these will control when shuffle pads thinks an "entire column of
    # pads" really is one. This allows it to be fooled into deleted columns that
    # are almost pads. It's definition is - calculate the consensus and if it's a
    # pad then it's an entire column. The consensus_mode 0 (frequency) and cutoff
    # 1 (100%) are the defaults, which therefore really do mean entirely pads.
    set_def CONTIG_EDITOR.SHUFFLE_PADS.CONSENSUS_MODE       0
    set_def CONTIG_EDITOR.SHUFFLE_PADS.CONSENSUS_CUTOFF     1

     
    • James Bonfield
      James Bonfield
      2006-01-03

      Indeed the replacement does not have this feature any more.

      However it turned out that rather a lot of people liked the contig editor shuffle pads feature purely for its "column-of-pads removal" code, despite having general universal distaste for the rather poor realignment ability.

      This wasn't something I'd realised before hand, but most recently I added it back as "Strip Pads" within the editor. This new version does no pad movement at all - it's purely for removing columns of pads. Due to the name change the new settings were also changed:

      set_def CONTIG_EDITOR.STRIP_PADS.CONSENSUS_MODE 0
      set_def CONTIG_EDITOR.STRIP_PADS.CONSENSUS_CUTOFF 1

      It's not in 1.6.0 though yet, although if you're brave it is in the CVS tree (src/gap4/edCommands.c would be a starting point).

      I plan on releasing 1.6.1 as a bug fix to 1.6.0 soon (for possibly long periods of "soon" if past history is anything to go by though) as there's a few other issues to sort out.

      James