FeatureCounts and E.coli Reference

[mitchum20]

Hi all,

I am struggling to count Bowtie2-mapped reads using featureCounts and after three days of investigation I hope you can help me.

My workflow is simple.
Bacterial (E.coli) Illumina RNAseq data -> FASTQC -> Bowtie2 to custom E.coli strain reference -> Samtools subsampling -> FeatureCounts.

Bowtie2 Mapped Reads: 34,445,043
FeatureCounts: 2,672,611 (7.8%) Successfully assigned alignments

The call:
featureCounts -a reference.gtf -p -T 10 -M -O -t CDS -g ID -o output.txt mapped.bam

Why does featureCounts only annotate 7% of the reads, even though the entire genome is full of perfectly annotated genes (see attachment) ?
How to run featureCounts properly on bacterial RNAseq data?
Is there an alternative software to featureCounts?

Please have a look at the IGV attachment. Everything looks fine, I am very desperate how to debug this situation. I tried almost all combinations of featureCounts paramters. I used PROKKA and BAKTA genome annotations.

If you have any guess or idea how to come closer to the origin of the problem please let me know.

Here is the featureCounts output:
Assigned 2672611
Unassigned_Unmapped 0
Unassigned_Read_Type 0
Unassigned_Singleton 0
Unassigned_MappingQuality 0
Unassigned_Chimera 0
Unassigned_FragmentLength 0
Unassigned_Duplicate 0
Unassigned_MultiMapping 0
Unassigned_Secondary 0
Unassigned_NonSplit 0
Unassigned_NoFeatures 31772432
Unassigned_Overlapping_Length 0
Unassigned_Ambiguity 0

Best,
Michael

[mitchum20]

I solved the problem. The issue was that most of the reads correspond to 23S rRNA which is only visible if you change the featureCounts type to rRNA.