How to explain the result of rMATS

[machun...@gmail.com]

Hi all and rMATS team,

I ran rMATS to detect the alternative splicing events for my RNA-seq data a few days ago. Even though I have already read the documents regarding the output of MATS in http://rnaseq-mats.sourceforge.net/user_guide.htm, I'm still confused about the meaning of each record in each file of MATS_outputs folder because I found that the gene Cox6b1 in the annotation file mm10/gencode.vM7.annotation.gtf only had one transcript in the SE pattern event result, I got two isoform ratio for this gene. 

The command that I ran the rMATS:

python /home/zhangbo/Tools/rMATS.3.0.9/RNASeq-MATS.py -b1 1Aligned.sorted.bam,2Aligned.sorted.bam,3Aligned.sorted.bam -b2 19Aligned.sorted.bam,20Aligned.sorted.bam,21Aligned.sorted.bam -gtf reference/mm10/gencode.vM7.annotation.gtf -o result -t single -len 50 -analysis U

note: I tried to compare two conditions: A group vs G group, 3 replicates with condition.

In the mm10/gencode.vM7.annotation.gtf, the gene Cox6bs has annotation as follow:

But in the SE pattern result, I got two records for this gene:

As you can see, the gene Cox6b1 in the annotation file mm10/gencode.vM7.annotation.gtf only has one transcript in the SE pattern event result, I got two isoform ratio for this gene in SE pattern result.  In addition, this gene appears in both SE pattern result and MXE pattern result. Since it only has one kind of transcript, how it can have two splicing events for this gene? Does anyone can help me explain it or tell me how to find a document that can explain the result in detail?

Any help will be appreciated.

Thanks in advance,

Michael

[Juw Won Park]

rMATS not only find known (or annotated) AS events, it also finds novel (or unannotated) AS events. Even though there is only one isoform for a gene, RNA-seq can have reads supporting more than one isoform and that's how rMATS detects novel AS events. 

-Juw Won