Closed JohnHadish closed 4 years ago
I raised this same issue here #68. I closed it because I was under the impression that GEMs without the RowID entry work now in both Python and R. However if R still generates the RowID then I can add a checkbox in the import-emx analytic to tell KINC to ignore the RowID.
This was a really easy change so I just went ahead and added it.
This is a minor suggestion that I think would solve unnecessary user confusion.
I am of the opinion that both of the following examples should be qualified as valid GEMs, and that KINC should automatically accept either:
GEM A
GEM B
The only difference between them is that 'GEM A' contains
geneID
in the first column while 'GEM B' does not. Currently, KINC only considers 'GEM B' a valid option. If 'GEM A' is submitted, an error similar to the following will be thrown:While this is an easy fix for the user (just delete the first word), it is rather annoying small thing that I think the user should not have to deal with.
I am therefore of the opinion that KINC should identify if the first row has either the same or -1 entries than the second row, and handle it accordingly. This would solve unnecessary headaches, making both of the above GEMs valid.
As I said before, this is a minor thing, but I think it should be addressed for a better user experience.
SUPPLEMENTAL The Annotation matrix requires the word "Sample" in row 1 col 1. This means that it's format is opposite of the GEM's format.