Are parent features lines always appear before sub-feature lines in GFF3 files?

Question

[bill-zt]

In GFF3 file, there are parent features lines and sub-feature lines like:

I   SGD gene    335 649 .   +   .   ID=gene:YAL069W;biotype=protein_coding; 
I   SGD mRNA    335 649 .   +   .   ID=transcript:YAL069W;Parent=gene:YAL069W;biotype=protein_coding

Usually, parent features lines (such as gene) should appear in front of sub-feature lines (such as mRNA). I'm not sure whether it is a custom or a strict rule. In the GFF3 Specifications document (https://github.com/The-Sequence-Ontology/Specifications/blob/master/gff3.md), there is no such rules.

Answer 1

[Michael]

As to my knowledge, there is no such requirement in the format definition of GFF itself anywhere. However, when you want to import GFF3 features into a database, e.g. using the CHADO schema, in an efficient way, the features need to be sorted like this. There is a perl script in the CHADO package, that sorts and groups gene models by position and parent relationship: gmod_gff3_preprocessor.pl