Inconsistent input files for update_gff

[Neato-Nick]

Hi, I'm correcting a query assembly based on a reference assembly, and updating the query gff for the corrections. But, I'm hitting the "Inconsistent input files" error.
Here's my command and the error from updategff:

ragtag.py updategff PHRA102.gff PR-102_v3.1_cor-PA-u/ragtag.correct.agp
Thu Sep 29 09:15:16 2022 --- VERSION: RagTag v2.1.0
Thu Sep 29 09:15:16 2022 --- CMD: ragtag.py updategff PHRA102.RXLRs.CRNs.gff PR-102_v3.1_cor-PA-u/ragtag.correct.agp
##gff-version 3

Phyram_PR-102_s0001	AUGUSTUS	gene	133386	134116	.	+	.	ID=PHRA102_1;Name=hprt1_2;locus_tag=KRP23_1
Traceback (most recent call last):
  File "/nfs5/BPP/Grunwald_Lab/home/carleson/opt/conda/envs/ragtag/bin/ragtag_update_gff.py", line 162, in <module>
    main()
  File "/nfs5/BPP/Grunwald_Lab/home/carleson/opt/conda/envs/ragtag/bin/ragtag_update_gff.py", line 156, in main
    sup_update(gff_file, agp_file)
  File "/nfs5/BPP/Grunwald_Lab/home/carleson/opt/conda/envs/ragtag/bin/ragtag_update_gff.py", line 114, in sup_update
    raise ValueError("Inconsistent input files.")
ValueError: Inconsistent input files.

Here was my command for correction:

ragtag.py correct -u --gff PHRA102.gff -o PR-102_v3.1_cor-PA-u $ref_assembly PR-102_v3.1.fasta -t 16

And here are the first 10 lines of my GFF:

##gff-version 3
Phyram_PR-102_s0001	AUGUSTUS	gene	133386	134116	.	+	.	ID=PHRA102_1;Name=hprt1_2;locus_tag=KRP23_1
Phyram_PR-102_s0001	AUGUSTUS	mRNA	133386	134116	.	+	.	ID=PHRA102_1.1;Parent=PHRA102_1;Dbxref=CDD:cd06223,InterPro:IPR000836,InterPro:IPR005904,PFAM:PF00156,TIGRFAM:TIGR01203,UniProtKB/Swiss-Prot:Q6WIT9;Name=hprt1_2;Ontology_term=GO:0009116,GO:0004422,GO:0006166;locus_tag=KRP23_1;product=Hypoxanthine-guanine phosphoribosyltransferase
Phyram_PR-102_s0001	AUGUSTUS	exon	133386	133391	.	+	.	ID=PHRA102_1.1-exon1;Parent=PHRA102_1.1;locus_tag=KRP23_1
Phyram_PR-102_s0001	AUGUSTUS	exon	133454	134116	.	+	.	ID=PHRA102_1.1-exon2;Parent=PHRA102_1.1;locus_tag=KRP23_1
Phyram_PR-102_s0001	AUGUSTUS	CDS	133386	133391	1	+	0	ID=PHRA102_1.1-cds1;Parent=PHRA102_1.1;locus_tag=KRP23_1;product=Hypoxanthine-guanine phosphoribosyltransferase
Phyram_PR-102_s0001	AUGUSTUS	CDS	133454	134116	1	+	0	ID=PHRA102_1.1-cds2;Parent=PHRA102_1.1;locus_tag=KRP23_1;product=Hypoxanthine-guanine phosphoribosyltransferase
Phyram_PR-102_s0001	AUGUSTUS	intron	133392	133453	.	+	.	ID=PHRA102_1.1-intron1;Parent=PHRA102_1.1;locus_tag=KRP23_1
Phyram_PR-102_s0001	AUGUSTUS	start_codon	133386	133388	.	+	0	ID=PHRA102_1.1-start_codon1;Parent=PHRA102_1.1;locus_tag=KRP23_1
Phyram_PR-102_s0001	AUGUSTUS	stop_codon	134114	134116	.	+	0	ID=PHRA102_1.1-stop_codon1;Parent=PHRA102_1.1;locus_tag=KRP23_1

And first 10 lines of the AGP:

## agp-version 2.1
# AGP created by RagTag v2.1.0
Phyram_PR-102_s0001	1	152786	1	W	Phyram_PR-102_s0001_1_152786_+	1	152786	+
Phyram_PR-102_s0001	152787	381574	2	W	Phyram_PR-102_s0001_152787_381574_+	1	228788	+
Phyram_PR-102_s0001	381575	422454	3	W	Phyram_PR-102_s0001_381575_422454_+	1	40880	+
Phyram_PR-102_s0001	422455	556621	4	W	Phyram_PR-102_s0001_422455_556621_+	1	134167	+
Phyram_PR-102_s0001	556622	711649	5	W	Phyram_PR-102_s0001_556622_711649_+	1	155028	+
Phyram_PR-102_s0001	711650	727677	6	W	Phyram_PR-102_s0001_711650_727677_+	1	16028	+
Phyram_PR-102_s0001	727678	800993	7	W	Phyram_PR-102_s0001_727678_800993_+	1	73316	+
Phyram_PR-102_s0001	800994	1033585	8	W	Phyram_PR-102_s0001_800994_1033585_+	1	232592	+

Based on the code for updategff I'm getting this error because column 6 of the AGP don't have a match in the GFF I provided, but I thought that I'm using this script for the exactly described purpose. Let me know if I'm reading the documentation wrong.
Thanks for any help!

[Neato-Nick]

I realized I needed to use the -c flag so the tool looks for the AGP object in the gff instead of looking for the AGP component.

Then I found that updategff requires no gaps are in the agp. Got around that error by adding a new flag -s for when splitasm was used, maintaining current functionality by allowing gaps only if -s was set.
I still don't have a complete fix because I ran into the issue that some of the genes in the gff overlap stretches of N that were removed by splitasm. I could either

• Change the ValueError to a warning and just drop any feature overlapping Ns
• Add a -gff option to splitasm just as implement in the correct module

[Neato-Nick]

Implemented fixes in #135