Seeing reads in both directions from a directional library?

[arctan]

I made bacterial RNA-seq libraries with ScriptSeq v2 and did a small sequencing test (MiSeq Nano, Single Read, 50 bp).

ScriptSeq v2 prepares directional libraries, as per the manual: "The sequence produced by the Read 1 Sequencing Primer is that of the sense strand of the original fragmented RNA molecule."

After trimming and QC, I aligned the reads against the reference genome using BowTie.

When I visualize the resulting SAM files on a genome viewer (IGV), I see reads overlapping annotated CDS features going in _both_ directions. This is definitely _not_ what I expected, since the library should be directional.

Since I am new to this, I am hoping to understand might be going on. Is there something wrong with my library?

Is it possible that there was some genomic DNA contamination in my RNA preps (although I did treat with DNase and I didn't see a large peak on Bioanalyzer) which then became fragmented and incorporated as a sequence-able molecule in the library?

What are the possible reasons for seeing bidirectional reads from what was supposed to be a directional library?

[kmcarr]

Not surprising for bacterial RNA-Seq. Bacteria will transcribe large tracks of their genome from both strands, in polycistronic primary transcripts which may overlap. The implication is that reads aligning to the reverse strand of an annotated CDS may represent the primary transcript of a nearby gene on the opposite strand downstream of the annotated gene you are looking at.

[arctan]

Thanks for the reply! I will keep that in mind and a quick literature search does show significant antisense transcription in bacteria.

How would I distinguish real antisense transcription from genomic DNA contamination in the RNA-seq reads?

[asanair]

Hi
I am Asha, We are planning to do Miseq WGS on food samples and completely new to this field. Our aim is to identify the food content , species specific to look for meat adulteration etc. which will be the a suitable pipeline for this, it will be great if you can put some inputs into this.
thanks
Asha

[arctan]

Well, I ended up getting the results of my run and the directionality (strand bias) is so bad and I reported about it here: http://seqanswers.com/forums/showpos...10&postcount=5

Note to Asha (asanair): I'm pretty new to this stuff. It's probably a good idea to post your new question on a new thread with a suitable subject line, instead of posting it as a reply in this thread. It's probably also better to be as specific as possible about your questions. Good luck! I know I certainly need it