featureCounts problem annotation

Question

[hsu]

I have download Saccharomyces cerevisiae (Yeast) genome and annotation from Ensembl R64-1-1.

code as：

featureCounts ./Saccharomyces_cerevisiae/Ensembl/R64-1-1/Annotation/Archives/archive-2015-07-17-14-36-40/Genes/genes.gtf -o genecounts_888 -t gene -p -g Name .exprnasamout/SRX3084888.bam

There is the mistake :

Failed to open the annotation file /datc/wangjc/Saccharomyces_cerevisiae/Ensembl/R64-1-1/Annotation/Archives/archive-2015-07-17-14-36-40/Genes/genes.gtf, or its format is incorrect, or it contains no 'gene' features.

Is there other Saccharomyces cerevisiae (Yeast) genome and annotation sources?

[ATpoint]

Give a picture of head -n 20 genes.gtf please. Also, input files (at least in the most recent version) need to be indicated by -a.

[hsu]

I       ensembl start_codon     538     540     .       +       0       exon_number "1"; gene_biotype "protein_coding"; gene_id "YAL068W-A"; gene_name "YAL068W-A"; gene_source "ensembl"; gene_version "1"; p_id "P5379"; transcript_biotype "protein_coding"; transcript_id "YAL068W-A"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS5441";

    I       ensembl transcript      538     792     .       +       .       gene_biotype "protein_coding"; gene_id "YAL068W-A"; gene_name "YAL068W-A"; gene_source "ensembl"; gene_version "1"; p_id "P5379"; transcript_biotype "protein_coding"; transcript_id "YAL068W-A"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS5441";
    I       ensembl stop_codon      647     649     .       +       0       exon_number "1"; gene_biotype "protein_coding"; gene_id "YAL069W"; gene_name "YAL069W"; gene_source "ensembl"; gene_version "1"; p_id "P3634"; transcript_biotype "protein_coding"; transcript_id "YAL069W"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS1129";

[ATpoint]

Ok, looks like a normal file. Use:

featureCounts -a genes.gtf -t 'exon' -o countMatrix.txt input.bam

Answer 1

[michael.ante]

Usually you count on the exon feature level, which is grouped on gene level. The gene feature which you are specifying with -t gene is not a standard feature (see here) and may be not included in your file.

Nevertheless, you'll get a table with the read count per gene if you leave the -t parameter to its default value.

[hsu]

I got these two files genecounts_888 genecounts_888.summary without -t gene

Is this correct ?

[ATpoint]

Yes that is correct because 'exon' is the default and that is the correct choice for standard RNA-seq.

[hsu]

Thank you very much

[ATpoint]

If the michael.antes answer solved your problem, please consider to mark it as accepted to help others in the future.