Counting reads supporting a specific splice junction from bam file
Hi,
How can I count the number of reads supporting a specific splice junction (e.g. chr1:15000-20000 where 15000 is the splice donor and 20000 is the splice acceptore position) in a RNA-Seq bam file aligned with STAR ?
Thanks
• 6,151 views
•
link
updated
by
Ram
•
written
by
Nicolas Rosewick
•
written
by
Nicolas Rosewick
1 answer
Log in to answer this question.
More posts like this
-
polyA imapct on downstrem alaysis particularly alternative splicing
written by mahsa80.mm •
-
Exon and junctions count per gene
written by rhonddaskl •
-
featureCounts exon counting problem
written by Nicolas Rosewick
-
Polyadenylation cleavage sites detection from rRNA-depleted RNA-Seq
written by Nicolas Rosewick
-
How to extract reads supporting protein-coding gene splice junctions from STAR's SJ.out.tab output
written by Nicolas Rosewick
-
Possible to use Oncofuse on the splice junction output file of rna-star?
written by Mark Pieterse •
-
Local Alignment Of Rna-Seq Data
written by Nicolas Rosewick
-
Removing Duplicate Reads Post Alignment
written by Nicolas Rosewick
-
Average Percentage Of Rna-Seq Reads Coming From Known Annotation (Ensembl, Refseq,...)
written by Nicolas Rosewick
-
Splice Junction Mapper For Dense Genome
written by Nicolas Rosewick
Doesn't STAR's .junc file provide that count?
In the
*.Log.final.outthe number of introns/splices are recorded but I am not sure if this this is the number of spliced reads. I have checked the manual but couldn't find an answer. A read can be spliced multiple times of course.Thanks Devon I forget about the SJ.out.tab file. However if someone do not have this file and only have the bam file. Any idea?
Ah, I'd forgotten the exact name, thanks.
If you don't have that then you'd have to follow Irsan's suggestion. Of course, this would likely mean needing to code something since I don't personally know anything that would do it. In theory, that should be possible with a bit of pysam to get the splice junctions. The simplest method would then be to write a script to output only splice junctions and then to pipe that to
sortand thenuniq -cfollowed byawkto reformat. The performance might not be optimal, but for an ad hoc solution I figure that'd work.