NexteraXT Bionalayzer - undertagmentation?

[Meyana]

Hi,

I am using the NexteraXT kit to prep mtDNA. I just did a test prep of some samples and had them run on a HS DNA Chip on the Bioanalyzer (see attachment).

I would like a second opinion on these. Does the libraries look undertagmentated?

(see past sample B, I screwed that one up...)
A: Following the procedure recommendations (1ng)
C: Half of the recommendations (0.5ng)
D: Quarter of the recommendations (0.25ng)

Also, quantifying the libraries by Qubit revealed them to be on the low end, just under 2nM with standard 12cycles PCR. Should I be concerned that the output of the prep is so low?

Thanks!

[Meyana]

Would still appreciate any input

[nucacidhunter]

Would you elaborate on the experiment aim because interpretation would depend on the final aim.

Edit: Are these post PCR libraries cleaned with beads?

[Meyana]

Would you elaborate on the experiment aim because interpretation would depend on the final aim.

Edit: Are these post PCR libraries cleaned with beads?

The aim of the experiment is to identify mtDNA mutations (heteroplasmy). We willbe doing paired-end seq on the MiSeq. We will do around 5000X coverage to enable this.
Before prep I did an "enzymatic purification" to remove nDNA and only retain mtDNA from gDNA purified from mouse tissue.

These are post-PCR, AMPure purified libraries. Eluted in volume recommended by the protocol (and then 1/2 and 1/4 for samples C and D respectively, also using 1/2 and 1/4 beads for cleanup).

[nucacidhunter]

I think that A is better because the other two show signs of low diversity library. For optimum read number you would need to quantify the library with qPCR and take the average size of area under 100-950 bp when calculating molar concentration.

You can change bead ratio to modify the size cut if you want to merge R1 and R2. If you do 2x300 cycles on current library A you probably will get around 30% merge.

[Meyana]

Could you please elaborate on how you determine which of the libraries have lower diversity? Is it due to the "spiky" pattern?
(I am new to prep of samples, previously I just passed them on to another person in the lab, but I would like to know how to best assess the libraries).

Also, inherently my samples are very low complexity. Based on some quality checks of my protocol, I will estimate I have >95% mtDNA (compared to nDNA) in my sample before going into library prep. The mtDNA is only ~16kb. So no matter what, wouldn't this be considered a low complexity sample? And then it would make sense if the libraries show low complexity?

I am doing paired-end because I will be looking for mtDNA deletions which are very frequent in my model mouse system and may be important in disease symptom development. So as such, I will not merge R1 and R2.
I was thinking to use the v2 2x150bp kit.

[nucacidhunter]

Spiky pattern is good indicator of ow diversity.

Your samples if tagmented randomly would produce around 15,000x15,000 fragments so it is not low diversity. PhiX is only 5.4 kb.

This library would be fine for sequencing.

[GA-J]

A and C looks good to me, D looks over-fragmented due to low starting mass.

B ? The size looks too large.

But I think you will have good data yield for A,C and D, may receive less data from B.

[Meyana]

Thanks for the input from both of you!

nucacidhunter
For some of my samples I will not be able to input more than 500ng (at the max, working with very small brain tissue pieces) to the prep, which is why I was attempting to scale down the reaction. Do you think my output will suffer tremendously if I am to sequence something that looks like C or D?

GA-J
What makes you say D is over-fragmentated? When I compare to the NexteraXT manual and look at the examples provided by Illumima I would say all my libraries are a bit large...

[nucacidhunter]

Thanks for the input from both of you!

nucacidhunter
For some of my samples I will not be able to input more than 500ng (at the max, working with very small brain tissue pieces) to the prep, which is why I was attempting to scale down the reaction. Do you think my output will suffer tremendously if I am to sequence something that looks like C or D?.

I guess you mean 500 pg and that should be fine. With your 16kb target and 5000x coverage you require 80Mb sequencing per sample. The required read number (2x150) would be ~267k plus some reads to cover off target regions.

D might be pushing it and you might get more duplicates but should not affect final results.

[Meyana]

Thank you for all your kind input! The libraries are coming along very well. Will get BioAnalyzer results from the last ones today (had to redo a few)

Now comes the next part, pooling the libraries.

My concentrations varies from 1nM to 5.6nM.
According to the protocol, I should pool equimolar concentrations to achieve a 2nM final pooled library.

However, since some of my samples below this concentration, how should I deal with this?

The protocol recommends using 5uL 2nM library with 5uL 0.2N NaOH. Could I instead use 10uL 1nM library with 1uL 1N NaOH (and disregard the 1 extra uL, since everything will be diluted with the addition of 990uL HT1).

Thanks!

[GA-J]

You may pool equal mol for each sample, then find the final pooled lib con. Meanwhile, if you receive 0.5nM, look at the protocol, you could do as below: Starting Library Concentration Library 0.2 N NaOH
4 nM 5 µl 5 µl
2 nM 10 µl 10 µl
1 nM 20 µl 20 µl
0.5 nM 40 µl 40 µ

The protocol: NextSeq System
Denature and Dilute Libraries Guide Document # 15048776 v02

[Meyana]

Thanks! - I have only looked at the MiSeq protocols (since I will be using the MiSeq), but I guess this is also applicable for the MiSeq?

[GA-J]

For Miseq, whatever you started, you may adjust to load 10pM. I have done this a few times, it worked well. You don't have to spike more than 5%, I used 1% spike.

[Meyana]

Great to know. I was just about to ask how to choose between 6, 8, and 10pM, I haven't seen any info or guidelines on how you know what you should do. Could you add any further information for my future runs? (this is mostly a method test)

I was thinking to do a 5-10% PhiX spike. Some of my libraries show low diversity, as adviced above in this thread (the 'spiky' pattern on the BioAnalyzer, only for a few of the samples).