Specific question about the strands in Illumina paired-end libraries

Question

[Diego D.]

I have a couple of questions regarding Illumina Sequencing. I have done some research on the subject, but I haven't been able to find a satisfactory answer.

Some preamble: When we generate paired-end libraries, I know that every read in a pair corresponds to the 5->3' end of one of the strands of the fragment.

Assuming that upper/lower is the ordering we defined for the original DNA molecule being sequenced. A pair of reads would be like this:

Fragment A:

upper strand
     R1
   ----->
5' ---------------------------- 3'
3' ---------------------------- 5'
                         <-----
                            R2
                       lower strand

In this case, R1 is placed in the forward FASTQ file and R2 is placed in the reverse FASTQ file.

Now, look at this other example:

Fragment B:

lower strand
     R1
   ----->
5' ---------------------------- 3'
3' ---------------------------- 5'
                         <-----
                            R2
                      upper strand

Note that in fragment B, reads are in the opposite direction, that is, R1 corresponds to the lower strand and R2 to the upper strand.

Is it possible to generate a pair of reads in the opposite direction? (as in Fragment B). If possible, why is that so? Is it because of the sequencing adapters? Is this the reason we have to consider both strands when aligning reads?

Thanks in advance!

[biofalconch]

Hi Diego,

Is it possible to generate a pair of reads in the opposite direction? (as in Fragment B). If possible, why is that so? Is it because of the sequencing adapters?

Yes, it is totally because of the sequencing adapters, remember that they are not specific so they will bind to either strand of the DNA thus generating opposing reads pairs

Is this the reason we have to consider both strands when aligning reads?

Pretty much, yeah

In RNA-seq, this becomes a bigger issue, and thats why there is strand specific sequencing.

[Diego D.]

Thanks for the quick response!

[GenoMax]

Diego D. : Take a look at document attached to post #2 in this SeqAnswers thread. That should be useful.

Aligners will automatically RC reads to check both strands.

[Diego D.]

Hi genomax,

Thank you for the feedback. I will check that document out!