#5 BED file for infer_experiment.py

[Tomas Pluskal]

Hi,
I have paired-end data from RNA-seq experiments on non-model species (no genome available). I would like to check if library was prepared as strand-specific. Can I use the infer_experiment.py script for that?

I can generate a de novo transcriptome using Trinity or rnaSPADES and align my reads to the transcriptome using bowtie2, that way I can generate a .bam file. But the infer_experiment.py script also requires a .bed file, how do I obtain that?

Thanks,

Tomas

[Liguo Wang]

RSeQC compares the "strand of reads" (after alignment) to the "strand of gene" (from your BED file) to determine if the RNA-seq experiment is strand-specific or not. Without reference genome and refernce gene model, RSeQC cann't tell if the library was prepared as strand-specific.

After de novo assembly, you basicaly get mRNA sequences (partial or complete), but you still don't know which strand (+ or -) the mRNA is encoded.

Liguo

[Tomas Pluskal]

Thanks for the response.

It is true that I don't know which strand the assembled transcripts belong to. However, if the library prep was strand-specific, then for each assembled transcript all F reads that map to it should map in the same direction, right? If it was unstranded, then the F reads would map equally in both directions. Or am I missing something?