Difference between GFF3 and GenePred Starting Base Location

[RogerRoger]

Hi all,

I am using gff3ToGenePred to convert a gff3 file and I noticed a small difference that occurs--all starting locations (txStart, cdsStart, exonStarts) are one less than the number found in the gff3 file.

So, for example, if I have only one CDS in a gene and it's listed in the gff3 as starting at 45000 then corresponding cdsStart in the GenePred file will be 44999.

It only does this for starting locations (not end locations), and it does it consistently, so I think it's more likely due to a difference in how GFF3 and GenePred interpret "where to start" (with GFF3 saying it starts "at this location" and GenePred saying it starts "immediately after this location")

Can anyone with more experience confirm this is the case or do I write a script to re-adjust starting position back to the value found in the GFF3?

Thanks!

[GenoMax]

GFF3 has a 1-based co-ordinate start where as GenePred has a 0-based start.

Official explanation at this link.