RSeQC infer_experiment strange output

Question

[nonolescargot]

Hello

I tried the tool RSeQC infer_experiment.py on my RNAseq bacterial data. I know that my data is reverse stranded but I get the following output :

This is PairEnd Data
Fraction of reads failed to determine: 0.0693
Fraction of reads explained by "1++,1--,2+-,2-+": 0.3577
Fraction of reads explained by "1+-,1-+,2++,2--": 0.5730

With the doc of RSeQC we would have expected that we obtain like 0.99 for the "1+-,1-+,2++,2--" because it is reverse stranded, or 0.5/0.5 if it wasn't stranded. But we obtain a strange middle ground.

Do you have any idea of what is happening with the data ?

Thank you

Answer 1

[Istvan Albert]

map your reads to the genome of interest and visualize what the alignments look like.

What clarifies a lot of ambiguities is if you align only the data from pair 2 as single-end data.

All the reads from pair 2 should align in the sense (or antisense, depending on the library prep) direction.

Now you can visualize which do not and why.