5', 3' SS and Intron retention read count using STAR out.tab

Question

[thindmarsmission]

Hi!

I am interested in the study of alternative splicing sites 5 and 3 and Intron retenton. I just want it, in the form of read count that how many numbers of reads and percentage of reads/total coverage reads, belongs to one splice site and same thing for other site.

NOTE: I don't have two different conditions.(so i am not considering MATS and related tools)

I aligned reads to reference using STAR and I have STAR out.tab file with unannotated and annotated events. Now, How I can extract information like listed below for 5',3' SS and IR events.

example [5' SS]:

                 start                    end                      number_of_reads
  chr1            10300                10500                      40
  chr1            10350                10500                       150

[Eric Lim]

You might be interested in tools like SGSeq: https://bioconductor.org/packages/release/bioc/html/SGSeq.html

It reads from the aligned bam. It isn't the fastest tool, but from personal experience, it's pretty sensitive for both annotated and novel events.

Note: 3'ss and 5'ss stand for 3' splice site and 5' splice site, respectively. AS stands for alternative splicing. People in the field don't use ASS an acronym.

[thindmarsmission]

Thanks for suggestion. I am just trying out this tool. Do you think, I can find Intron retention events using this tool?

[Eric Lim]

Yes, my best understanding is that most tools rely on annotated exons to call intron retention (IR). Some tools infer IR by accessing exon-intron junction reads at both 5'ss and 3'ss. This approach is generally correct but will produce some false positives on those introns when both splice sites are alternatively spliced.

[Malcolm.Cook]

No: based on my reading of the paper and skimming the source code, it appears to me that SGSeq does not analyze IR.