Bowtie and % of RNA-Seq mappability.

Question

[tarek.mohamed]

Hi All,

I am analyzing RNAseq data for 12 samples with and without treatment performed on Hiseq illumina platform (paired end,100 bp reads, 40 million reads / sample) quality of fastaq files is fine. At this step, I am interested in DGE rather than splicing data.

I used Bowtie integrated in QuasR package for mapping the RNAseq data to reference genome (GRCh38).

The % of reads mapped to reference genome is about 50% in all files, is this % ok to perform DGE

Thanks

Answer 1

[Sukhi Singh]

Generally, you don't high mappability when mapping RNA-Seq reads to the sequenced genome because it's a special methodology in sense, that you reverse transcribe the RNA to DNA and sequence which miss intronic regions. That's why special tools like Tuxedo suite, remap the reads that did not mapped the first time to special libraries based on exon-exon junctions and later both the files get combined and this is what you use to call DEG using multiple replicates and samples.

QuasR seems to general short read mapping and processing tool. Use a specialised tool, I believe you will find nice RNA-Seq reviews on BioStars itself.

Good Luck!