What to do when degrated genomic DNA is the only option for ddRAD library prep?

[apfuentes]

Hello everyone,

I am in the situation where degraded DNA is the only option for ddRAD library preparation (picture attached showing 5 uL of total genomic DNA per well in a 0.8% agarose gel-0.5X TBE, run at 62 V for 45 min).

I wonder what other researchers have done/recommend to face this issue. I am thinking on these strategies:

1. Clean up selecting for DNA with high molecular weight.
- l found on the internet the Genomic DNA Clean and Concentrator kit (Zymo Research. Have any of you used it before?

Or,

2. Cutting the band with the largest molecular weight from an agarose gel.
- I found Qiagen QIAEX II Gel Extraction kit (DNA size limit: 40 bp to 50 kb) and Zymoclean™ Large Fragment DNA Recovery Kit (DNA size limit: 50 bp to ~200 kb) for this purpose. Maybe the Zymo's kit would be better because of the larger fragment recovery limit, have you used any of these ones?

Thank you in advance for any recommendation.

[nucacidhunter]

l found on the internet the Genomic DNA Clean and Concentrator kit (Zymo Research. Have any of you used it before?

I have used it but that is not going to remove smaller fragments. It will clean up DNA and remove inhibitors.

Cutting the band with the largest molecular weight from an agarose gel.

That may not make any difference for your ddRAD and you will loos around 50% of DNA. The loss with Zymo kit is higher in comparison to QIAEX II.

My suggestion would be not to do any of above. Degradation not only has affected the size of fragments but it also would cause other damages not seen on the gel such as nicks, chemically altered nucleotides and … ,which will affect downstream reactions. One optional step is to use DNA repair enzyme mix to fix damaged DNA before proceeding to library prep. In RADseq methods equal contribution of samples to pooled digested-ligated restriction fragments is ensured by equal input. In this case because of damage and degradation this assumption will not be met. So, if you have limited number of samples it would be best to use 1-2 ug input into your digestion and ligation reactions and skip the size selection. After PCR, quantify individual libraries and pool equal amount of each and do post-PCR size-selection. That should give good results (at least in theory). If variation in samples tags or coverage is high, you can do more sequencing to obtain adequate coverage for all tags.

[apfuentes]

Thanks nucacidhunter for your great advice!

Two questions:
1. Is there any DNA repair enzymatic mix you could recommend me please over the others?

I found 2 types so far:
- One is used only for blunting and phosphorylation of DNA ends,
Thermo: http://www.thermoscientificbio.com/m...nd-repair-kit/, or
NEB: https://www.neb.com/products/e6050-n...-repair-module

- And the other type fix a lot of other prePCR things,
NEB: https://www.neb.com/products/m0309-precr-repair-mix


2. Your suggestion of DNA repair, skipping size selection after digestion and adapter ligation, do PCR, and better quantify individual libraries after PCR sounds good to me, because I would be enriching for fragments that have P1 and P2 adapters, ignoring the undesired background fragments caused by degradation.

What I think could be problematic would be pooling equal amount of each and do post-PCR size-selection. Different DNA samples may have different proportions of P1-P2-fragments vs. degradation-caused fragments within the size range of 300-400 bp. So, even if I equal the amount of DNA for pooling (based on PicoGreen dye measurements) and then do size selection, in the pooling step I would probably have different amount of P1-P2-fragments among individuals. How could I deal with that? I think that quantification with Kapa quanti library kit could give me a more precise estimation of the amount of P1-P2-fragments per individual library although this would increase a lot the cost of the project (n=800). Any ideas?

[nucacidhunter]

1- PreCR Repair Mix is the most suitable product and a similar product is used in preparing PacBio long templates.

2-

So, even if I equal the amount of DNA for pooling (based on PicoGreen dye measurements) and then do size selection, in the pooling step I would probably have different amount of P1-P2-fragments among individuals. How could I deal with that?

Pooling by qPCR may not give better results than mass pooling because, it will quantitate all sizes of fragments up to a size that extension time allows and not the size range you are interested in. To deal with uneven coverage among samples, after first sequence run, under-represented libraries can be pooled again and size-selected for extra sequencing. That is the main reason for doing individual PCR rather than pool PCR. If PCR amplicons are going to be size-selected, it is important to note that size-selection window of 300-400 bp will translate to 250-350 bp equivalent of pre-PCR size. In this method it would be best to carry out bead clean-up of ligation reactions with 1x AMPure beads rather than 1.5x. That is to remove non-target small fragments which will be preferentially amplified.

I think that quantification with Kapa quanti library kit could give me a more precise estimation of the amount of P1-P2-fragments per individual library although this would increase a lot the cost of the project (n=800). Any ideas?

With 800 samples (if it is not number of qPCR reactions with replicates), the best option might be to pool a portion of digestion-ligation reaction by volume and follow standard protocol. Remaining reactions can be frozen. After sequencing, again under-represented samples can be pooled and run through protocol and sequenced.


If you are using NEB restriction enzymes that can be heat inactivated you may also skip the clean-up of digestion reaction which has been suggested in ddRAD paper and after heat kill go straight to ligation by addition of a ligation mixture containing ligation buffer, ligase, adapters and supplementary ATP for the volume of digestion reaction (1 mM final). This is especially very useful if the adapter ligation destroys the restriction enzymes recognition sites. I do not think that heating DNA fragments over 100 bp to 65°C would denature it as it has been suggested in the paper.