High Amount of UnAssignedNoFeatures in FeatureCounts Output

Question

[Tom Brady]

I have some bulk (Illumina) RNA seq paired end data. I aligned it to the genome with STAR, turned the .sam to .bam and sorted with samtools. I tried to quantify with featureCounts, but am getting an high amount of Unassigned_NoFeatures:

Here's the featureCounts command I'm using in R:

fc <- featureCounts(
  files = sorted_bam_file,
  annot.ext = annotation_file,
  isGTFAnnotationFile = TRUE, 
  GTF.featureType = "gene", 
  GTF.attrType = "ID",  
  useMetaFeatures = TRUE,  
  allowMultiOverlap = FALSE,  
  isPairedEnd = TRUE,  
  requireBothEndsMapped = TRUE, 
  nthreads = 4,  
  strandSpecific = 0  
)

P.S. The annotation is in GFF

[rfran010]

What library type is it (polyA selected, totalRNA + rRNA-depletion...)?

What annotation file are you using in what organism?

One potential example could be if rRNA is not (well) depleted and rRNA locations are not in your GTF file, then you might expect such a large percentage of reads are outside your features. It will take some digging to know what is going on.