Library Prep with Low Input

[Joltex]

It sounds like a number of people have been able to prepare successful libraries for sequencing from as little as a few ng of starting DNA. Does anyone have any experience preparing libraries from such small amounts and/or specific tips to increase the chances of success?

I've read some suggestions such as modifying adapter concentrations, increasing the number of PCR amplification cycles and performing PCR before size selection. It seems like there's a lot more information out there for Illumina than for Ion Torrent, including kits such as the Nextera and NEXTflex kits.

Thanks!

[MrGuy]

Looks like you aren't looking hard enough:
library prep -- http://www.neb.com/nebecomm/products/productE6270.asp
"Nextera" equivalent -- http://products.invitrogen.com/ivgn/...search-product

Regardless, I am always looking for the lowest amount of input DNA. A lot of the papers seem to rely on Qiagen Repli-G or other MDA techniques. I am confident bias is introduced with the methods as I have also noted the normalization and averaging of the genomes by pooling with the WGA methodology.

Here's one using 454:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0018565

[jtackney]

I guess that all depends on what kind of libraries you want - full genome, exome, amplicon. You can always do an initial shotgun library and then enrich. As for only finding information for Illumina, I don't think that should be a limiting factor - whatever someone learns to optimize library creation on that platform can easily be ported over to Ion Torrent (now if you want to talk about through-put differences, you have a point/problem). The emPCR vs cluster generation shouldn't be an issue once you get to an amplified library.

If you are concerned about losing material in the cleanup steps, I've been tempted to go down the route that Broad is using and doing a "with-bead SPRI" and only eluting at the very end (http://www.ncbi.nlm.nih.gov/pubmed/21205303).

If you are concerned about blunt-end ligation with A-B adapters, convert to Y-adaptors for 454/Ion Torrent (http://www.ncbi.nlm.nih.gov/pubmed/21886102 , http://www.ncbi.nlm.nih.gov/pubmed/20435675).

If you wanted more ideas on working with small amounts of DNA, I would look into the ancient DNA literature. There they are either working with very small amounts of DNA OR small amounts of true target DNA in a sea of lots of contaminating soil DNA.

I would also look into the ChIP-DNA literature.

To speak to your specific comments, yes, start off with dropping adapter concentration a ton. Use qPCR on unamplified library aliquots to determine cycles to right-before-plateau and amplify to that. If you need to go above ~20 cycles, do two different amplification steps. Many are switching over to SPRI from column purification (apart from automation, one big reason is to better eliminate adapter-dimers ), though you might lose more DNA that way. I agree with MrGuy that WGA certainly introduces Biases and I personally don't touch it for NGS (though others have certainly been successful).

[Joltex]

Thanks for all the information! It's very helpful.

I've been having to prepare a number of different libraries - amplicons, transcriptomes, and WGA product from single cells, - but all with significantly less than the recommended 100 ng input. I should note that I'm new to the Ion Torrent and NGS in general so I don't have much experience.

Both the cleanup and size-selection steps concern me when starting with so little. We've been using bead purification but the "with-bead SPRI" method you posted sounds really interesting. Let me know how it goes if you try it.

This may be a silly question but is there a particular rule-of-thumb that you follow when adjusting adapter concentration? Life doesn't seem to specify the concentration in their kits but, as an example, if the recommended volume of adapter to use with 100ng of DNA is 2uL would you simply scale that to 0.5uL for 25ng of DNA?

[IonTorrent]

Hi Joltex,

Regarding the adaptor concentration in the kit- for the kits supported by the current Ion Xpress Plus Fragment Library Preparation User Guide, the adaptors are supplied at a concentration of 20 uM A adaptor and 20 uM P1 adaptor.

Thanks for all the information! It's very helpful.

I've been having to prepare a number of different libraries - amplicons, transcriptomes, and WGA product from single cells, - but all with significantly less than the recommended 100 ng input. I should note that I'm new to the Ion Torrent and NGS in general so I don't have much experience.

Both the cleanup and size-selection steps concern me when starting with so little. We've been using bead purification but the "with-bead SPRI" method you posted sounds really interesting. Let me know how it goes if you try it.

This may be a silly question but is there a particular rule-of-thumb that you follow when adjusting adapter concentration? Life doesn't seem to specify the concentration in their kits but, as an example, if the recommended volume of adapter to use with 100ng of DNA is 2uL would you simply scale that to 0.5uL for 25ng of DNA?

[jtackney]

The Ion adapters therefore are suppose to be used at a working concentration of .4 uM (for 50-100ng). I have gone down to .04 uM each adapter working.

And personally, I wouldn't try Meyer's single-stranded method as your first go around with low quantity DNA. It's really meant to recover low quality DNA - where one strand has a polymerase blocking lesion. I'm sure it works fine for low quantity in general, but I think it would be over kill (and a drastic change in your workflow).

Finally, I wouldn't gel size select. You will certainly lose too much there as well.