Are the mate-pair reads always shadowed by their friends the paired-end (picture)

[seb567]

** Title should be "Are the mate-pair reads always shadowed by their friends the paired-end reads (picture)"

So, are the Illumina mate-pair reads always shadowed by their friends the paired-end reads ?

See picture -> http://imgur.com/LNwws

In the picture above, I can see two populations:

- 250 +/- 25
- 2250 +/- 225

(approximations)

What cause these during the library preparation ?

I thought only the 2250 +/- 225 would be fished out by the biotin/streptavidin step.


(data are from Assemblathon 2, this is the snake dataset)


Thanks !

Sébastien
http://twitter.com/sebhtml

[tonybolger]

So, are the Illumina mate-pair reads always shadowed by their friends the paired-end reads ?

In short, it seems yes.

In the picture above, I can see two populations:

- 250 +/- 25
- 2250 +/- 225

(approximations)

What cause these during the library preparation ?

I thought only the 2250 +/- 225 would be fished out by the biotin/streptavidin step.

Indeed - those fragments which span the splice are enriched, relative to those which come from a linear stretch of the original DNA. But the latter massively out-weigh the former (increasingly so at longer MP lengths - since there are more bad fragments, still only one good fragment), so the strep merely tilts the balance a bit in favour of the 'from the splice' fragments. The ratio can go from something like 10:1 against the spiced fragments to maybe 5:1 in favour - which isn't all that bad.