htseq-count error Feature does not contain a 'gene_id' attribute

Question

[madzayasodara]

I'm trying to do read count using htseq-count

htseq-count --mode=union --stranded=no --idattr=gene_id -f sam ./filteredfirstZBND1X_gff.sam ./filteredsecZBND1X_gff.sam /home/madzays/data/finc h_data/ZF_transcriptome/taeGut2_gff/taeGut2.gff > ZBND1X_gff_counts.sam

but get the error

Error occured when processing GFF file (line 11 of file  /home/madzays/data/finch_data/ZF_transcriptome/taeGut2_gff/taeGut2.gff):
Feature id1 does not contain a 'gene_id' attribute   [Exception type:
ValueError, raised in count.py:77]

this is what my gff file (from NCBI) looks like

##gff-version 3
#!gff-spec-version 1.21
#!processor NCBI annotwriter
#!genome-build Taeniopygia_guttata-3.2.4
#!genome-build-accession NCBI_Assembly:GCF_000151805.1
##sequence-region NC_011462.1 1 118548696
##species http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=59729
NC_011462.1     RefSeq  region  1       118548696       .       +       .       ID=id0;Dbxref=taxon:59729;Name=1;chromosome=1;gbkey=Src;genome=chromosome;isolate=Black17;mol_type=genomic DNA;sex=male
NC_011462.1     Gnomon  gene    17853   28371   .       +       .       ID=gene0;Dbxref=GeneID:100227449;Name=CLIC6;gbkey=Gene;gene=CLIC6;gene_biotype=protein_coding;partial=true;start_range=.,17853
NC_011462.1     Gnomon  mRNA    17853   28371   .       +       .       ID=rna0;Parent=gene0;Dbxref=GeneID:100227449,Genbank:XM_002186596.2;Name=XM_002186596.2;gbkey=mRNA;gene=CLIC6;partial=true;product=chloride intracellular channel 6;start_range=.,17853;transcript_id=XM_002186596.2

[cpad0112]

1. Try to convert gff to gtf and feed it to htseq-count. try searching for gffread util.
2. Shift to featurecounts.

Answer 1

[WouterDeCoster]

Yes. There is no gene_id attribute in your gff. The error is pretty self-explanatory in that sense.

Try with --idattr=gene. Based on the few lines you posted that attribute is present.

[madzayasodara]

K it's doing its thing now. thanks!

[madzayasodara]

Oh wait it run into a line without that attribute. Do you know if NCBI gffs have any "must have" attribute I could use? thanks

[WouterDeCoster]

Nope. You can print out the line which caused the problem using e.g.
sed -n 26p taeGut2.gff for line 26, then try to figure what that line does have in common with the rest.

[madzayasodara]

Yep it runs for "gbkey". However it outputs

-bash-4.2$ head -50 ZBND1X_gff_counts_reverse.sam
C_region        0       0
V_segment       0       0
exon    0       0
mRNA    0       0
misc_RNA        0       0
ncRNA   0       0
precursor_RNA   0       0
rRNA    0       0
tRNA    0       0
__no_feature    21536704        15508244
__ambiguous     0       0
__too_low_aQual 0       0
__not_aligned   6233570 4564356
__alignment_not_unique  2340806 1668760

I think it has to do with the gtf file having these RefSeq ids instead of chromosome locations (e.g. chr1) Any ideas how could I sub them?

[WouterDeCoster]

Perhaps ensembl has a gff for your organism?