Directional Illumina read mapping for prokaryotic transcriptomes

[Matt.Ralston]

Hello SeqAnswers!
I am looking for mapping software to analyze directional/strand-specific RNA-seq data from both paired-end and single-end libraries of bacterial transcriptomes. The libraries are prepared with either the TruSeq and ScripSeq library prep kits.
I have considered using TopHat for the analysis, since I am aware of its --library-type fr-firststrand option to preserve the directional information. However, I am concerned that its default settings may not be optimal for prokaryotic transcriptomes.
Downstream, I would like to identify transcriptional start sites (T.S.S.) and antisense RNAs (asRNAs).

1. Is TopHat optimal for analyzing directional RNA-seq data from bacteria?
2. Do you know of a thread or information about a TopHat parameter set that is suited for prokaryotes?
3. Are any other alignment algorithms capable of analyzing directional data?


Thank you SeqAnswers and Happy New Year!

[swbarnes2]

All aligners will tell you whether you are looking at read 1 or read 2, and whether the read is reversed with reference to the sequence, or not.

Tophat is just Bowtie, but it tries to align across splice sites. But prokaryotes don't really have splice sites, do they?

I don't think you need specialized software here. Just use whatever aligner you usually use.

[dsher]

Hi Matt,

Try Rockhopper - http://cs.wellesley.edu/~btjaden/Rockhopper/index.html . We have been using it and it seems quite user friendly. The newest version knows how to deal with stranded info.

TopHat is not very good at identifying genes in bacteria, especially if they are densely coded or even overlapping. Another potential program to use that was developed for prokaryotes is EDGE-Pro (Magoc et al, Evolutionary Bioinformatics 9 (127-136), 2013).

BTW, I would be interested to hear what your experience is in looking at the antisense reads. We seem to have quite a lot of them, and are wondering what is noise and what is real.

Daniel