Definition clarifications by @phillipeloher

[lpantano]

Not sure if opening an issue here is appropriate or if there is a better place to put these questions. No real issues to report in this case, just some clarification/documentation questions.

header:

what is 'source ontology', example?
for 'sample names used in attribute' - does this include items such as SRA accessions?
'small RNA GFF version ' - is that the format of this specification? Or is this a different version?
Filter tags, really like the concept. Think we should give an example that's not just Pass/Fail. Such as CAUTION(ambiguous sequence, may not be an isomiR)
Attributes:

Read Name - is this optional when the attribute->expression flags are used?
Parent - is this precursor name the same as Column 1 precursor name? If so, is it redundant?
Hits - what is defined as a 'hit'? Is this different than Attributes->Expression?
Variant - what happens it multiple variants for the same isomiR? Will open up a different issue on this since it's more involved

[lpantano]

• we have examples here for : https://github.com/miRTop/mirGFF3/blob/master/definition.md, if the database is in a official resource then we use that, if not the URL of the database. That was something the parent GFF format had, but still is not implemented.

• for sample names, is whatever the user wants to use actually, we can add more information for sure.

• yeah, I would love to add official TAGS: I'll open an issue for this to see what people offer.

• Read name, is optional, you can add it for debugging reasons

• Parent can be redundant if the tool mapped against precursor, but if it was mapped to genomics, or translate to genomic coordinates at some point, then is the chromosome name.

• Hits means to say the number of time the read is present in the GFF, for instance for ambigous reads.

• Thanks for opening a new issue with the variant issue!