Hey guys,
I have 8 samples, each sample having 4 lanes. How do I go about mapping the sample to my reference when there's multiple lanes?
I'm not sure how cating the 4 lanes together would work out for counts. Here's my thought process:
If I cat the 4 lanes together before mapping (BWA), and then run BWA mem, won't that give me 4x the expression/counts since the mapping consists of 4 duplicate reads?
If this is actually what would happen, or do I have the process all wrong?
How would you properly map/align these 4-lane reads?
1 answer
Splitting a library over 4 lanes doesn't result in 4 copies of every read, no one would spend the extra money to run the additional 3 lanes then. Just concatenate the files and map them, though I wouldn't recommend bwa for RNAseq (it's not splice aware, try STAR or hisat or even BBMap, which quite flexible).
Log in to answer this question.