Hello,
I am using featureCount to complete a gene level count of my RNAseq reads. From the results summary, it looks like I have a very high number multimapping reads, resulting in a low assingment rate (less than 15%).
I have paired end reads, with an unstranded library
The options I have chosen are as follows:
featureCounts -T 8 -p -a ~/file/path/sacCer3.ncbiRefSeq.gtf -t exon -g gene_id -o ~/file/path/fcresults/groseq2.txt UWN.bam UMN.bam
after processing the reads, this is what my summary looks like:
Status UWN.bam UMN.bam
Assigned 7608607 10093518
Unassigned_Unmapped 4743940 3388412
Unassigned_Read_Type 0 0
Unassigned_Singleton 0 0
Unassigned_MappingQuality 0 0
Unassigned_Chimera 0 0
Unassigned_FragmentLength 0 0
Unassigned_Duplicate 0 0
Unassigned_MultiMapping 66981505 58121996
Unassigned_Secondary 0 0
Unassigned_NonSplit 0 0
Unassigned_NoFeatures 1927159 2654380
Unassigned_Overlapping_Length 0 0
Unassigned_Ambiguity 10983810 8513544
From what I understand, this means I have reads (fragments) that are mapping to more than one feature.
My question is, if the SAM/BAM file I created was able to successfully map fragments to a reference genome, how is it possible that these fragments are assigned to more than one gene (meta feature) in my GTF file? Do I need to use a different annotation file, or is this a typical assignment rate for the count I am after?
Any help would appreciated! Thanks,
1 answer
From what I understand, this means I have reads (fragments) that are mapping to more than one feature.
I think the 67M read are aligning to multiple places in the genome, and another 11M are aligning to multiple features.
That's not normal. In the RNA preps I analyze, I typically get STAR telling me that about 70% of reads mapped to the genome uniquely.
Here are some stats from a recent experiment; Total are all the reads in the fastqs, mapped means they were aligned to the genome, assigned means they were assigned to a gene (I used RSEM for the last step, it is smarter about handling reads that are ambiguous that featureCounts is)
Assigned_reads Mapped_reads Total_reads
Sample_1 22858584 26435275 27173101
Sample_2 15934143 18462699 18732273
Sample_3 20612644 23995078 24387303
These might be a little nicer than usual; the RNA was commercially-made control RNA.
Log in to answer this question.
-
Low "assigned alignments" in output and high "Unassigned_NoFeatures" in summary
written by 2406691063 •
-
Very low successfully assigned alignments with feature counts
written by Manuel •
-
featureCounts results: low rate of 'Successfully assigned alignments'
written by Hamtaro •
-
bam_stat.py and featureCounts stats don't seem to make sense (to me)
written by bertb •
-
Why there are low assigned reads in featureCounts output and different outputs for quality checks?
written by newbie
-
Using featureCounts on paired-end reads aligned with HISAT2 resulting in 0 assigned fragments
written by turo.1 •
-
No aligned reads after featurecounts. BAM files have >85% reads mapped
written by LAB_2019 •
-
Problem with extracting read counts from RNA seq file
written by popescuiofelia •
-
featureCounts: 0% successfully assigned fragments on multiple PE .BAM files
written by imesi •
-
featureCounts duplicate counting
written by lkmklsmn
