Overrepresented sequence in RNA-seq only in R2

Question

[benjamin.saintpierre]

Hi everyone,

After running some RNA-seq experiment, we have observed an over-represented sequence which do not correspond to any adapter used. It clearly seems to be a contaminant.

But my main question is why do I get this sequence only in one strand?

As I get the sequence in 2 400 000 reads, I don't get its reverse-complemented sequence. I don't know why, because if I get a contaminant sequence, I should get half of the reads with the sequence, and half with its complement, shouldn't I?

[WouterDeCoster]

Have you tried blasting it?

[benjamin.saintpierre]

Yes, it's 60S Rrna human, so a contamination is obvious. But what i can't explain is why is there only this: R2adapter - Sequence of 300 bases - R1adapter and none of this: R1adapter - Sequence of 300 bases - R2adapter The adapters must be set in a random way, no?

[Manvendra Singh]

May be first you map your reads against rRNA and tRNA sequences (preferably via bowtie), and discard the mapable reads. If remaining reads still show over-represented sequences then there is something weird, it could also be the case of chimeric transcripts.

[benjamin.saintpierre]

Actually, i've already done this mapping, and i know it's form this sequence: Chain 5, Structure Of The H. Sapiens 60s Rrna, 5070 bp. The problem is not what the sequence is, but why is it represented only in the 5'-3' strand, and not in the other way. I mean, if adapters are setting randomly around the rna fragment, why do i only observe R2adapter-Sequence of 300 bases-R1adapter, and not R1adapter-Sequence of 300 bases-R2adapter?

[WouterDeCoster]

Which library prep method was used for the RNA-seq?

[benjamin.saintpierre]

Script Seq, Illumina

Thanks for your answers/time!

[WouterDeCoster]

This kit generates highly directional RNA-Seq libraries from rRNA-depleted or poly(A)+ RNA from any animal, plant, or bacterial species. Ribo-Zero ribosomal RNA (rRNA) removal kits are recommended for use with ScriptSeq v2 to achieve whole-transcriptome, diverse, stranded libraries.

So it's strand specific and rRNAs are supposed to be removed.

[benjamin.saintpierre]

So if it is strand specific, it means that I will observe always get the R2adapter in 5' of my sequence and never in 3' ? So this: R1adapter-Sequence of 300 bases-R2adapter can't exist ?

[WouterDeCoster]

Yes, from the FAQ of Script-seq:

Does the ScriptSeq process generate directional libraries? Yes. ScriptSeq kits make directional libraries such that the sequence generated using the Illumina Read 1 sequencing primer is that of the sense strand of the original RNA. Using the Illumina Read 1 sequencing primer, the first nucleotide read always corresponds to the 5´ nucleotide of the original fragmented RNA molecule.

How does the ScriptSeq method generate directional libraries? After rRNA depletion (e.g., using Ribo-Zero Kits), the RNA is reverse-transcribed using random hexamers that include a 5´ tail containing a unique tagging sequence. After cDNA synthesis, a terminal tagging oligonucleotide (TTO) that has a blocked 3´ end is annealed to the 3´ end of the cDNA. Following extension with a DNA polymerase, a second, unique tagging sequence (complementary to the TTO) is added to the 3´ end of the cDNA. The resulting di-tagged cDNA is amplified by PCR using primers that anneal to the two different tagging sequences on the cDNA.

[H.Hasani]

You should not have RXadapter - Sequence of 300 bases - RYadapter in first place, as the first adapter will be sequenced by the primer anyway. Have you checked with the technician who did the experiment what went wrong? something seems very fishy on wet-lab level!

Answer 1

[benjamin.saintpierre]

Thnaks for all your answers! It's the strand specific experiment information that i didn't have in the first time.

Thanks for your time!