kbeutel
commented
10 months ago From Nathan: To test, make some .meta files with:
the same name for different chr/position/ref/alt (preferably with the same position, just different alts) -> should throw an exception that you need to manually change the names to be unique, but will be supported when that happens
the same name for the same chr/position/ref/alt (accidental duplicated line; we can't support this)
different names for the same chr/position/ref/alt (problem we can't support)
From Abby: Currently, if there are variants listed in a .meta file that have the same position, and both variants exist in the dataset, only one will be returned and included in the score, even when minWinSize=0. It would be nice to have an option (perhaps a minWinSize=-1 or something similar) where every variant in .meta that exists in the dataset would be included (I.e. no variants would be left out simply because of the same or a nearby position). This would be helpful for instances where a score generated from variants all within a gene/similar location and overlap/close proximity of variants is likely.