If you align your data with STAR aligner, it can output read per gene, and it does so three different ways,; assuming the protocol is unstranded, assuming your reads should align in the forward direction, assuming they should align in the reverse direction. So you can see which column has the most reads aligning best.
Hi,
If the column corresponding to the "reverse" (fr-firststrand) has very low number of reads assigned, and the "pos" (fr-secondstrand) column has a number of reads comparable to the "non stranded" column, can I assume/conclude the experiment is "pos" (i.e. stranded but the opposite way as for the standard Illumina TruSeq RNA-seq protocol)?
Here are the counts of reads assigned I have for a couple of samples (single end RNA-seq) to illustrate this:
sample reverse yes no
sample1 88,171 3,585,710 3,561,601
sample2 77,863 3,143,859 3,119,066
Thanks!
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
RSeQC infer-experiment-py is pretty good for this purpose, just go ahead and use it. QoRTs also can determine strand-specificity, but
infer-experiment.py
is more straightforward to use.