[sshell]

Adjustments for low input

Haven't needed to do this since I routinely make illumina libraries from 20ng or so starting material, and have been getting good sequencing results.

Jasonwood, can you tell me what modifications (if any) you make to your library construction protocol when starting with low inputs such as 20 ng? Do you use less adapter, for example?

Thanks so much!

[jasonwood]

I dilute the adapters down 1:20 (from 50uM stock concentration). I will usually increase the PCR cycles from 15 to 17-18, especially for ChIP samples.
I use Ampure XP beads for all purification steps, as I've found them less lossy than QIagen columns for low amounts of DNA. Also I size select a pretty wide band (200-500bp).

[lterhune]

Jasonwood, are you using the TruSeq kit to prepare the ChIP libraries or the Paired-End kit? We had lousy results with our first round of TruSeq ChIP libraries, but the protocol is not yet optimized.

We are also trying to create a library with about 200ng of gDNA- we have both the TruSeq and Paired-End kits, and our own reagents for our "homemade" Paired-End protocol.

Thanks!

[jasonwood]

I don't use Illumina kits, everything is homebrew. For Truseq adapters, the conditions are a little different than what I outlined above. They self-ligate much more readily than the older style short adapters, so I dilute them down more. For 20ng ChIP input, I do 1:50 at least. And change the Ampure bead ratio from 1.8:1 to 1:1 for the cleanup steps after the ligation and PCR enrichment (this effectively depletes fragments shorter than 200bp). Finally, I started measuring input to the PCR enrichment, and I put 2ng DNA in for 10 cycles. This gives you plenty of product and ensures you don't run out of PCR primers. If the starting amount is especially low (e.g. 5ng) or the post-ligation yield is very low, I will sometimes boost the cycle number, but generally now I use fewer cycles than I used to. I've had great results with this method, and minimal adapter contamination in my reads.

[lterhune]

Awesome, thank you .

We are also trying to use as few PCR cycles as possible because we having issues with high duplication rates. I haven't tried limiting the input ligation for the PCR enrichment before...

[lterhune]

Do you do a size-selection step with your ChIP libraries after PCR, or eliminate it completely?

[jasonwood]

2ng and 10 cycles works very well, for both ChIP and RNA. I elute after the ligation step into 10ul and Qubit it, and that way I can usually get a reading. If it is too low to read I just put the whole thing in.

[jasonwood]

I size select after PCR on 2% agarose. However, I usually don't size select before the ligation, so I just do it once. Probably one way or the other is ok. I like to see what I am cutting out, and make sure the PCR enrichment didn't introduce a large MW smear, so I choose to just do it once at the end, and that works well for me.

[djones]

New product

Has anyone tried this yet? It looks like it's a product from Sigma designed for whole genome amplification prior to sequencing. http://www.sigmaaldrich.com/life-sci...ification.html

I wonder if it works well.

[claumer]

WGA prior to Illumina was something I was interested in too, but I think it would be very challenging if the purpose was de novo assembly, as it *will* cause coverage bias, and most de novo genome assembly algorithms out there I think assume roughly even coverage.

This could be a useful paper:

http://www.ncbi.nlm.nih.gov/pubmed/21441209

What WGA approach were you looking at in particular?