gene counts with gff3

[Marie Strader]

Hi, 

For some reason I am getting zero counts for genes using a gff3. The manual says a gff file is fine and I have noted this in my command, however it still ends up with an error and no counts. I would appreciate any advice on this!

Here is a preview of the gff3 file:

chr10 maker gene 78574 86389 . + . ID=Amillepora16770;Name=Amillepora16770;Alias=maker-chr10-snap-gene-0.16;

chr10 maker mRNA 78574 86389 . + . ID=Amillepora16770-RA;Parent=Amillepora16770;Name=Amillepora16770-RA;Alias=maker-chr10-snap-gene-0.16-mRNA-1;_AED=0.13;_QI=51|0.5|1|1|1|1|3|427|72;_eAED=0.13;

chr10 maker exon 78574 78653 . + . ID=Amillepora16770-RA:exon:0;Parent=Amillepora16770-RA;

chr10 maker exon 84594 84712 . + . ID=Amillepora16770-RA:exon:1;Parent=Amillepora16770-RA;

chr10 maker exon 85892 86389 . + . ID=Amillepora16770-RA:exon:2;Parent=Amillepora16770-RA;

chr10 maker five_prime_UTR 78574 78624 . + . ID=Amillepora16770-RA:five_prime_utr;Parent=Amillepora16770-RA;

chr10 maker CDS 78625 78653 . + 0 ID=Amillepora16770-RA:cds;Parent=Amillepora16770-RA;

chr10 maker CDS 84594 84712 . + 1 ID=Amillepora16770-RA:cds;Parent=Amillepora16770-RA;

chr10 maker CDS 85892 85962 . + 2 ID=Amillepora16770-RA:cds;Parent=Amillepora16770-RA;

chr10 maker three_prime_UTR 85963 86389 . + . ID=Amillepora16770-RA:three_prime_utr;Parent=Amillepora16770-RA;

My featurecounts command:

featureCounts -F gff -a Amil.coding.gff3 -exon -g Parent -o geneCounts.txt 100-1_S180.trim.sam

In the output file, I get the following results; it appears to identify the metafeatures but does not assign any reads:

...

Warning: Unknown annotation format: gff. GTF format is used.

...

Load annotation file Amil.co ... ||

||    Features : 190939                                                       ||

||    Meta-features : 28188                                                   ||

||    Chromosomes/contigs : 588                                               ||

||                                                                            ||

|| Process SAM file 100-1_S180.trim.sam...                                    ||

||    Single-end reads are included.                                          ||

||    Assign reads to features...                                             ||

||    Total reads : 984667                                                    ||

||    Successfully assigned reads : 0 (0.0%)                                  ||

||    Running time : 0.13 minutes                            

Sam files look like: 

D00289:62:CDR44ANXX:7:1101:3264:2171 0 Amillepora23113-RA 1503 255 44M * 0 0 AGACACAAGTTTTCAATTTCCTGTCAAGGAGAAAATTGATCACG FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF AS:i:74 XN:i:0 XM:i:2 XO:i:0 XG:i:0 NM:i:2 MD:Z:23C3A16 YT:Z:UU

D00289:62:CDR44ANXX:7:1101:19665:2101 0 Amillepora08292-RA 4434 0 15S20M9S * 0 0 ACGTGTGCTCTTCCGATTTTTTTTTTTTTTTTTTTTTTTTTTTT FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF AS:i:40 XS:i:40 XN:i:0 XM:i:0 XO:i:0 XG:i:NM:i:0 MD:Z:20 YT:Z:UU

[Yang LIAO]

The chromosome names in your GTF file is like chr10, but the chromosome names in your BAM file are like Amillepora08292-RA.

I see that your GTF file has a tag in the extra info column named "ID", and its value is something like "Amillepora16770-RA". However, featureCounts need to have the chromosome names in the GTF file (the first column) match the chromosome names in your BAM files (the 3rd column).