Orientation of paired-end reads

[marnilush]

Dear Forum,

being relatively new to NGS data analysis i have a newbie question: I got paired end solexa data and want to assemble it. To do so using assemblers like e.g. abyss, i need to specify the read orientations. I was assuming that this would be forward for the first reads and reverse for the second reads - i am not sure, though and googling did not really yield an answer. Do paired end reads always have the same orientations or is this depending on the individual experiment?

Thanks in advance!

[kmcarr]

The standard Illumina paired-end protocol produces reads oriented pointing toward each other, just like good old fashioned Sanger paired reads, but the insert size is much shorter. Imagine a 300bp DNA fragment with ends arbitrarily labeled A & B. Read 1 and Read 2 would represent the ends of these fragments oriented inward.

A---------------------------------B
----->                       <-----
read 1                       read 2

[marnilush]

thanks for your quick reply!

cheers

[lmf_bill]

so, it means Read 1(+) Read 2 (-) or Read 1(-) and Read 2 (+), and the genomic positions of positive read is less than negative read

[coldturkey]

what about solexa mate-pairs e.g. 3kb ?

Are they orientated --->...<---- as well or are they orientated <----...---->.

If the latter do they boh need to be reverse complemented before assembly or is that assembler specific

regards

Brian

[kmcarr]

so, it means Read 1(+) Read 2 (-) or Read 1(-) and Read 2 (+), and the genomic positions of positive read is less than negative read

Yes, that is correct Imf_bill.

what about solexa mate-pairs e.g. 3kb ?

Are they orientated --->...<---- as well or are they orientated <----...---->.

If the latter do they boh need to be reverse complemented before assembly or is that assembler specific

regards

Brian

Brian, the reads from the mate-pair protocol are oriented outward as indicated in your second drawing. I know that Bowtie has an option to specify the relative orientation of the pairs. The default behavior is appropriate for Illumina short insert paired-ends (inward pointing reads). I don't know about other aligners.

[JackieBadger]

It seems our MiSeq mate paired data isn't orientated as described above.

In a sequence that would be
A-------------B
We see

A----- B-----

Additionally each FASTQ file is mixed with both mate pair ends.
Is there a tool which will reverse one of these fastq files to get
A----- -----B

??

Thanks,
J