featureCount Too Many Genes

Question

[gtasource]

Howdy folks,

I just finished up a pipeline that involved differential expression of some RNA-Seq mouse files. Using the latest refseq assembly, I aligned the files, and used featureCount to generate counts (with the plan of using this output in DeSeq2.)

Anyway, I've been doing some checking out of the output file from featureCount, and when I looked at the number of rows in the file, there were close to 40,000.

Here's my featureCount command for starters:

featureCounts -t exon -g gene -a GCF_000001635.26_GRCm38.p6_genomic.gtf -o counts.txt -M bam1 bam2 bam3 bam4

So I'm curious. Are these extra genes redundant? Are there some pseudogenes in the list? What accounts for the large amount of genes.

Thanks!

Answer 1

[swbarnes2]

The latest ensembl mouse version shows 22K coding genes, and 16K non-coding genes, that makes close to 40,000 genes.

http://uswest.ensembl.org/Mus_musculus/Info/Annotation

[gtasource]

Thanks, after I sorted and calculated the number of unique genes, I realized there were no duplicates. Should've checked the annotation info first though. :P

Answer 2

[WouterDeCoster]

40000 genes is not wrong, it's perfectly normal. It also includes lncRNAs, for example.