Strand specific pair end Illumina sequencing

Question

[Bioinfonext]

I have strand specific pair end data. But I am not sure which option to use in trinity RF or FR, moreover I also do not understand which one left or right, I have a R1 and R2 read file for a library.

To find the strandness of pair end data I run infer_experiment.py script from RSeQC. IT gives me output like this:

**output result of RSeQC:

This is PairEnd Data

Fraction of reads failed to determine: 0.0196

Fraction of reads explained by "1++,1--,2+-,2-+": 0.9360

Fraction of reads explained by "1+-,1-+,2++,2--": 0.0444**

But I am not able to understand it. They have given this explanation for the infer_experiment.py script:

For pair-end RNA-seq, there are two different ways to strand reads (such as Illumina ScriptSeq protocol):

1.1++,1–,2+-,2-+

read1 mapped to ‘+’ strand indicates parental gene on ‘+’ strand

read1 mapped to ‘-‘ strand indicates parental gene on ‘-‘ strand

read2 mapped to ‘+’ strand indicates parental gene on ‘-‘ strand

read2 mapped to ‘-‘ strand indicates parental gene on ‘+’ strand

1. 1+-,1-+,2++,2–

read1 mapped to ‘+’ strand indicates parental gene on ‘-‘ strand

read1 mapped to ‘-‘ strand indicates parental gene on ‘+’ strand

read2 mapped to ‘+’ strand indicates parental gene on ‘+’ strand

read2 mapped to ‘-‘ strand indicates parental gene on ‘-‘ strand

Thanks in advance.

[zhenyisong]

I think you used the second strand to construct the library due to to the inference from the output by RSeQC. The second strand here means that the sequence order is in line with the mRNA and not the complementary DNA (cDNA). Sense-strand. If this is the case, you should use the FR. But you should check the definition for both FR and RF in the trinity setting. @Tom_L, the illumina stranded LT kit, the protocol generates the first strand to synthesize the DNA library(cDNA). Second strand was degraded using enzyme because of the usage of dUTP.

PS. from HISAT2: TopHat has a similar option, --library-type option, where fr-firststrand corresponds to R and RF; fr-secondstrand corresponds to F and FR

[Bioinfonext]

Hi Zhenyisong,

We use this kit for library preparation:

Ovation RNA-Seq Systems 1–16 for Model Organisms PART NOS. 0351 ARABIDOPSIS kit

Can you please confirm me again.

Thanks a lot

[zhenyisong]

I read the protocol here. for Ovation RNA-Seq Systems 1–16 for Model Organisms. The figure 1 at PDF page 6 said that this kit uses the first strand to construct the library. (Second Strand Synthesis with Nucleotide Analog, Ligation with Nucleotide Analog-marked Adaptors). But Please contact Nugen Specialist to confirm it. If this is true, this is in conflict with your RSeQC report. If you reverse the order of R1 and R2, this may result in this way?

[Bioinfonext]

Hi Zhenyisong,

Now according to library protocol what are you trying to suggest about library type: should I use RF or FR?

or should I do assembly without mentioning the type of library?

Thanks

[Bioinfonext]

I want to inform you one more thing: during first strand synthesis oligo dt and Random hexamer both were used.....

Does it also affect the strand specificity?

Answer 1

[Tom_L]

A strand-specific library from illumina sequencing certainly refers to dUTP sequencing. According to the RSeQC documentation, your results confirm that you deal with dUTP sequencing: http://rseqc.sourceforge.net/#infer-experiment-py

dUTP sequencing corresponds to the RF value in Trinity documentation: https://github.com/trinityrnaseq/trinityrnaseq/wiki/Running-Trinity

You can find information about strand specificity at: https://www.ncbi.nlm.nih.gov/pubmed/21943893 and http://likit.github.io/running-bowtiebowtie2-rsem-and-tophat-on-dutp-strand-specific-reads.html

Cheers.

[Bioinfonext]

Hi Thanks.....

Can you also confirm which one should be left and right in trinity (R1 and R2)?

R1=? R2=?

[Tom_L]

R1 is left, R2 is right.