Stringtie and featurecounts results: counts_gene.txt has more rows than counts_transcript.txt

[CHENAO-QIAN]

Description of the bug

I am running the pipeline with stringtie and featurecounts option. In stringtie2/featureCounts/, I got the counts table for genes and transcripts. However, counts_gene.txt has more rows than counts_transcript.txt. In counts_gene.txt, one MSTRG (stringtie gene) has several rows. I am wondering if counts_gene.txt is actually reporting exon counts.

Checking the featurecounts code in the pipeline, how were the parameters chosen/optimized?

featureCounts \\
        -L \\
        -O \\
        -f \\
        -g gene_id \\
        -t exon \\
        -T $task.cpus \\
        -a $gtf \\
        -o counts_gene.txt \\
        $bams

    featureCounts \\
        -L \\
        -O \\
        -f \\
        --primary \\
        --fraction \\
        -F GTF \\
        -g transcript_id \\
        -t transcript \\
        --extraAttributes gene_id \\
        -T $task.cpus \\
        -a $gtf \\
        -o counts_transcript.txt \\
        $bams

Thanks!

Command used and terminal output

Nextflow run nf-core/nanoseq -r 3.1.0 -profile singularity --input samples.csv --protocol cDNA --skip_demultiplexing --aligner mininap2 --quantification_method stringtie2 --skip_fusion_analysis --skip_differential_analysis 


No error.

Relevant files

No response

System information

Nextflow 23.04.1
HPC
Cmd
Singularity
CentOS
nanoseq 3.1.0