Target Sequence Enrichment 454

[dottomarco]

Hi everybody !
I am having a huge issue with the sequence enrichment for 454. My interest (and with me I think many of you folks) is to target small portions of the human genome and sequence many samples for the same target. Most labs out there do not have the money or simply do not care about sequencing whole exoms with the 454 but may be very interested in sequencing some hundreds of kbs or Mbs related to the genes they are studying.
The only available approved protocols for 454 sequence enrichment have been developed by Nimblegen which now offers
- Custom arrays (5Mb : 1000 € if you have the hybridator) with a minimum order of 5 arrays with the same probes design
- Exome arrays (30+Mb: 1400 € if you have the hybridator) without minimum of order
- Exome SeqCap EZ (liquid enrichment: like the SureSelect by Agilent) : (30+ Mb: 1000 €) - minimum order of 4 kits.
For us, 454 users there's NO custom liquid enrichment out yet. And more important : NO official protocol for multiplexing tagged samples with small target sequence (for ex. 5 samples with 1Mb target on the 5Mb Nimblegen array)
Agilent was the first provider of liquid (and so automatizable) sequence enrichment with its SureSelect which is available in 2 formats:
- Custom kit (3-5Mb target): 900 € each, with a minimum order of 10 kits with the same probes design (not really a good news for those who do not have 10 or more samples to enrich).
- Exome kit (30+Mb target)
They have only developed protocols for Illumina and SOLiD but nothing for 454. Neither is there any available news for multiplexing (a good cost-saving strategy could be to order a custom kit with a target of, for ex. 1 Mb and so a 3-5 X tilling, and use 1/3 - 1/5 of the solution to enrich each of the 3-5 samples).
The only other solution comes from Febit which enriches 1Mb in 2 weeks at 1300 € and is available for multiplexing up to 8 samples per chip (~150kb of target sequence each) for <500€ per sample. But this process requires sending them the sample and having to wait for the enriched sequences to come back each time we want to complete the process, which does not allow for in-house multiplexing and automation. Plus, Febit works only with Illumina and SOLiD.
I am asking if anyone in the SeqAnswer community could provide any feedback on this subject, or if anyone might have some updated news from those of you testing any of the following:
- FEBITcustom enrichment (multiplexed or not) with a 454
- Agilent liquid custom enrichment (multiplexed or not) with a 454
- Nimblegen arrays with tagged samples in multiplex
- Nimblegen SeqCap Ez custom (if any of you have had the chance to use their custom made kits which are not out on the market yet).

I thank in advance whoever can provide information regarding these issues.

[flxlex]

For us, 454 users there's NO custom liquid enrichment out yet. And more important : NO official protocol for multiplexing tagged samples with small target sequence (for ex. 5 samples with 1Mb target on the 5Mb Nimblegen array)

We once just did a (not-supported-by-Roche) MID tagging of two nimblegen seqcapture samples, and it worked very well. These were tagged at the stage of 454 library prep, and was done with the standard protocol (not the new 'rapid-library prep' protocol). So, you could try to find a sequencing service provide willing to try it out...

[flxlex]

Update (to prevent confusion):

The MID tagging was done after hybridization and amplification. The array design, hybridization etc was done in another lab. When we (i.e. the sequencing facility) got the samples, they were enriched, amplified and only needed to be prepared for the 454 sequencing, i.e. add adapters to them.

[dottomarco]

Thanks flxlex for the answer. With the present Nimblegen protocol you use the 454 adapters which are present in the library, to perform a PCR amplification after the enrichment. 1-2 years ago Nimblegen had their own adapters for the amplification ... so I guess your test was performed on samples which followed the old protocol with Nimblegen's extra adaptors, otherwise they would have prepared the library before the enrichment.

What I am looking for though, is any example of multiplexing the single enrichment procedure on Nimblegen's arrays or any test on Agilent or Febit enrichment systems followed by 454 sequencing. The idea is to enrich multiple small (100kb-2Mb for example) targets in order to lower the enrichment costs which now are almost as high as the sequencing itself. (Nimblegen for example charges 1000€ to enrich 5 Mb and the correspondent sequencing cost with a 30X coverage is not far from this. The enrichment of 1 Mb costs the same amount of € compared to 1/5 in sequencing costs)

Thanks

[nicorascovan]

Hi,

About a solution for your multiplexing enquires, I don´t have any suggestion. Beyond that, we are having issues also with the sequence capture. Besides the high price, another problem we find is the few options available for designing your capture arrays (mostly human and sometimes just a few other species). After looking for solutions on the web, we end up finding a company called Mycroarray (http://www.mycroarray.com/index.html). Does any of you know anything about them? They seem to give the option of designing truly full custom arrays. And the sizes don´t look that big as in the other technologies you´ve mentioned plus the price looks affordable. The Question is, are they suitable to complement with 454 sequencing? Can you trust them?

If any of you know anything about this, I´ll be happy to have your feedback.

Thanks.

[MYselect]

Hi,

Since Nicorascovan mentioned our company MYcroarray.com as a source for custom sequence capture kits for targeted sequencing, I would like to give you some details.

We have developed a capture kit called MYselect. It is fully customizable and we will work with you to design your probes. It consists on biotinylated 120mer RNA baits that you hybridize yo your genomic library. Baits and captured preys are pulled out of solution with magnetic beads and you can release your captured DNA by specifically degrading the RNA baits. Very similar to some of our competitors. The way we differ is in the probe synthesis technology.

For now, we offer kits containing 25,000 baits (I agree, its is small) for as little as $15 per reaction for 500 reactions up to $100 per reactions for a 5 reactions kit. Yes, it is $500 for a custom kit!. Soon, we will upgrade to 100,000 baits per kit. Our website (http://www.mycroarray.com/products/capture_arrays.html) has not yet been updated, but we will present these products at the Xgen conference in San Diego, March 17 - 19 2010.

Yes, we are fully suitable for 454 sequencing. We did some experiments in a collaboration with a beta testing site and got 60 to 80% reads on target on 4 different libraries, some MID-tagged. We ship our kit with a protocol for 454 libraries.

Please contact us ([email protected]) if you have any questions.
Thanks.

[nicorascovan]

Hi,

Thanks for your reply. I think this option for custom sequence capture is very useful and we might contact your company soon.

[NaMar]

sequence capture

Hi nicorascovan,

Have you test sequence capture with MYcroarray?
Thanks
Nathalie

[nicorascovan]

Hi Nathalie,

We haven't tried it yet, but we are planning to do it soon, end of this year, beginning of next.
If you try it before, let me know about your results.
Regards,
Nicolas.

[dottomarco]

We tried it. got good results about pretty must everything apart from % of reads on target (which is the actual important thing). we tested different tiling (2x, 5x, 10x and 2x multiplexing 2 samples during the enrichment) each test was conducted on 2 separate samples that got similar results. the best performer was the 2X test , followed by the 2X multiplexed.
The only problem is that, in the best case, we reached ~22/25% of mapped reads on target.
That's pretty poor, especially if compared w Sureselect (the only other custom enrichment kit available now on the market that states ~60/70% on target).
MYcroarray.com said they never got less than 50% on target... and that probably our target (~60kbp) was too small if compared with the others that they dealed with before and this may ve introduced some problems.
Only 2% of the target region was not covered at all (0X coverage - deep) if we consider all the reads produced by the lane (1/8 ptp used to sequence the 6 samples (+2 samples enriched in multiplex) derived from 2 genomic DNAs).
We are now trying to optimize the enrichment protocol and trying their new probe design (during the summer they ve introduced a new algorithm for the probe design) to see if we can get something better than 1/4 of the probes on target.

All the tests were conducted on a 454 (rapid libraries (MID tagged) - titanium)

[Zaag]

If you want to multiplex samples for a small region with a liquid sequence capture you could just aliquot the probes and hybridize your samples seperately.

[dottomarco]

yes, this was a test to try the "multiplexing" process in general... for small targets that's obviously not necessary.

Any idea how the seq enrichment perform with Sureselect for small targets (tens of kbs)?

doing some math on our experiment 20% on target (of a target region that represent 0.002% of the genome) is actually a 10,000 fold enrichment ... that is not bad.. I am wondering what we can get with a bigger target (1 or + Mbs)..

[MYselect]

With some larger targeted regions (2Mb), we got over 80% reads on target using the MYbaits kit (formerly MYselect). http://www.mycroarray.com/mybaits/MY...hment+kit.html
Please keep in mind that smaller the targeted region (or percentage of the genome to be targeted), smaller the number of molecule to be recovered. Assuming that the noise of non-specific sequence pulled down remain constant, then your percentage of on target reads will decrease with small targeted regions.
Decreasing the bait concentration when short regions are targeted seems to help increasing the percentage of on target reads.