Change-O modifications to support 10X data

Issue #147 resolved
Jason Vander Heiden created an issue
  1. Add definitions for LOCUS, C_CALL and CELL to the standard, with mappings to the appropriate AIRR Rearrangement fields (locus, c_call, cell_id)
  2. Add parsing of LOCUS to change.IO.IMGTReader and change.IO.IHMMuneReader. Already complete for change.IO.IgBLASTReader.
  3. Formalize paired cloning procedure. Ideally, as a separate subcommand to DefineClones that uses the CELL and FUNCTIONAL annotations to intelligently subset sequences for VH clustering, with either pre- or post-processing of clonal assignments using VL information.
  4. CreateGermlines will also need a solution for how to generate consensus germlines for cloned and paired files.

Comments (8)

  1. Jason Vander Heiden reporter

    The --10x argument to MakeDb is overriding the IgBLAST LOCUS call. If we want to keep the 10X chain annotation, we should probably put it in a different field (eg, LOCUS_10X). Though, the 10X call use the correct nomenclature (eg, IGH) and the IgBLAST calls don't (eg, VH)...

  2. Roy Jiang
    1. Looks you already did that.

    I would overwrite with IgBLAST. Seems easier to implement. But 2 needs to be done before we do this.

    3/4. Steve wants to wait for more discussion on this.

    Looks like you assigned to yourself so let me know if you need anything else from me.

  3. ssnn

    @Jason Vander Heiden we talked about this issue in the last meeting. We think the only point left is number 2:

    Add parsing of LOCUS to change.IO.IMGTReader and change.IO.IHMMuneReader. Already complete for change.IO.IgBLASTReader

    If this is correct, can we create a new issue for 2, and resolve this one?

  4. Jason Vander Heiden reporter

    Looks like I took care of the locus thing about a year ago (42c0d6c).

    If the plan is to deprecate DefineClones in favor of scoper, then we can just close this.

  5. Log in to comment