Extracting unconventionally spliced transcripts from RNASeq dataset

Question

[VenGeno]

Hi,

I am wondering what is the best way to identify and quantify an unconventionally spliced transcript (which is not included under the reference assembly) from a RNASeq dataset. This splicing is happening under special circumstances (). I already have the BAM files after reference-based alignment using HISAT2.

Thank you in advance.

[rpolicastro]

The easiest way would be to rerun HISAT2 with the --novel-splicesite-outfile option, which will output putative splice sites not in the reference file. Another option is running STAR in 2-pass mode, whereby novel splice sites are detected in pass 1, and then the alignment is rerun including those novel junctions in pass 2. I personally recommend this method.

[VenGeno]

Hi rpolicastro , Thank you for the reply. I ran HISAT2 on both modes (for something else; trying to see how many novels vs already identified) :). So, I have a novel splice site out files for all of my samples. I will check them to see if such is found in the region where my four alleles are located. If it is there how can I proceed to quantify the isoform? I was thinking one possibility may be to run stringtie in reference guided mode? Then do a isoform analysis using a tool such as IsoformSwitchAnalyzeR. Is there another way to assess? Thank you.