Comprehensive Review of Genomic Enrichment Technologies for Next Gen Sequencing

[ECO]

User sci_guy has posted a really great overview of genomic enrichment technologies in this thread: http://seqanswers.com/forums/showthread.php?t=8

He sent me his original powerpoint which contains 22 slides of well organized, detailed summary points on the current state of the art in targeted sequencing enrichment technologies. It is attached to this post.

Enjoy, and please comment here or in the specific relevant forum!



[Apologies to anyone catching a duplicate on the RSS feed...had some time issues, so I reposted this.]

[seqer]

In Okou et al, the enrichment is ~1000 fold, but only 1 in 60 reads are in targeted region. In Albert et al, the enrichment is 153-418 fold, but 38-76% of reads are in targeted region.

Could anybody comment why the percentage of reads mapped to targeted region is so low in Okou et al, but ok in Albert et al?

Thanks!

[Michael L. Altshuler]

I wonder whether anybody has attempted to reverse the enrichment strategy, that is, to capture the repeats and let exons be washed off the microarray. The washed material could be expected to be enriched in exons and be further processed for sequencing.

For all I know and can guess it could be a robust, efficient and inexpensive way of enrichment.

[sci_guy]

Interesting thought Michael.

The hyb kinetics would change a great deal. For example if you designed a series of oligos around the human LINE1 repeat consensus sequence you are now trying to capture over 20% of the genome (LINE1 sequences) with your oligos instead of capturing a megabase or two of exons (~0.05% of the genome). A liquid phase capture using a small library of oligos bound to magnetic beads or suchlike which are designed around repeat consensus sequences might be the way to go.

[apfejes]

That is interesting. For Solexa based applications, current alignment strategies don't handle repeats wel, and fragments aligning to repeats are most often thrown out - so the only significant advantage to removing these sequences might be in freeing up some additional capacity of a flow cell, eg, if you go from ~65% to 90% mappability of CFU.

It might be worth a try, if that's the case.

[Michael L. Altshuler]

Density gradient centrifugation would move most of the repeats to the satellite DNA peak(s).
As an initial step in the enrichment process it could probably alleviate the needle-in-the-hay situation during the subsequent array-based enrichment for the unique DNA sequences or even replace the arrays. Opinions?

[mcraja]

We are using Agilent's SureSelect on array capture method and "reverse enrichment" is 50% of what actualy happens, the repeat regions are removed by the use of Cot1 DNA in the hyb buffer. The target regions are captured on the array, there is more than 1000 fold enrichment of targted regions and > 99% of the targted regions are typically captured and more than 50% of the sequence is from targeted exons.

Raja
(will be posting updates at www.nextgenseq.com)

Originally Posted by Michael L. Altshuler

I wonder whether anybody has attempted to reverse the enrichment strategy, that is, to capture the repeats and let exons be washed off the microarray. The washed material could be expected to be enriched in exons and be further processed for sequencing.

For all I know and can guess it could be a robust, efficient and inexpensive way of enrichment.