Can Maker deal with a "dispersed_repeat" one-level features GFF3 file provided to the rm_gff option in the control file?
We're trying to annotate a genome of a non-model Drosophila species using a previously generated GFF3 file with repeats for this species. Some months ago, Carson Holt replied to a post (
https://groups.google.com/g/maker-devel/c/BhBMTF8dze8/m/y-QwJYRFAQAJ)
saying that "They (repeat models) must be match/match_part two level feature for rm_gff". However, all features on our GFF3 file are 'dispersed_repeat' one-level features.
Here is how our file looks like (extract of the file for a scaffold attached as well, repeats.gff3):
##sequence-region scaffold1062|size32755 1 32755
scaffold1062|size32755 RepeatMasker dispersed_repeat 1 47 271 - . Target=DNAREP1_DM 26 73;ID=2740
scaffold1062|size32755 RepeatMasker dispersed_repeat 48 83 263 + . Target=Dbuz_kz3_5_36225 704 739;ID=2741
scaffold1062|size32755 RepeatMasker dispersed_repeat 340 386 299 - . Target=rnd-1_family-30 19 65;ID=2742
scaffold1062|size32755 RepeatMasker dispersed_repeat 349 387 247 + . Target=rnd-1_family-29 1 71;ID=2743
scaffold1062|size32755 RepeatMasker dispersed_repeat 388 430 240 + . Target=rnd-5_family-3333_TIR_P 3669 3713;ID=2744
scaffold1062|size32755 RepeatMasker dispersed_repeat 446 555 408 - . Target=rnd-1_family-15 16 138;ID=2745
I'm attaching both the log file and the CTL file for what it's worth (Dtest.log / maker_opts.ctl).
I wonder whether Maker handled right our rm_gff file, whether the generated output is reliable. If not, do you think modifying the GFF3 file manually duplicating each row and changing the third column to 'match/match_part' could do the trick?
Thank you very much in advance.
Locked
Dtest.log (53K)