Hello. I am interested in using featureCounts as a tool for detecting abnormal splicing events in my RNA-seq data. I have a sample with a
known mutation which lead to exon skipping and I would have liked to count how many reads span this junction. I am trying now to decrease the number of artifacts i.e junctions that are supported by reads with low mapping quality. For that purpose, I add -Q flag to the command, but I
get not what I expected.
My command is:
featureCounts
-J -Q 255 -G GRCh38.primary_assembly.genome.fa -a
gencode.v32.primary_assembly.annotation.gtf sample.sam -o sample_fc
A one line example from output sample_fc.jcount:
PrimaryGene
SecondaryGenes Site1_chr Site1_location Site1_strand
Site2_chr Site2_location Site2_strand sample.sam
ENSG00000108821.13 NA chr17 50185429 + chr17 50186473 + 3308
ENSG00000108821.13 NA chr17 50185429 NA chr17 50194770 NA 1
ENSG00000108821.13 NA chr17 50185429 + chr17 50298247 + 1
ENSG00000108821.13 NA chr17 50185429 + chr17 50399843 + 1
ENSG00000108821.13 NA chr17 50185429 + chr17 50425454 + 1
The same command whithout -Q flag outputs the same result.
But when I use samtools view before this with -q 255 flag I got this:
PrimaryGene
SecondaryGenes Site1_chr Site1_location Site1_strand
Site2_chr Site2_location Site2_strand sample.sam
ENSG00000108821.13 NA chr17 50185429 + chr17 50186473 + 4
It
is obvious that reads are not filtered out for junction counting with
featureCounts, thus I wonder if it is not supposed to or I did something
incorrectly.
Thank you in advance for your response. A subsample of my sample is in the attachment in case needed