Size selection for Nextera libraries?

[sweetph3]

Hi. We have seen that for Nextera libraries, the smaller insert fragments cluster more efficiently than those with larger inserts. We ran our first 2x300 and most of the data is duplicated sequence as the inserts were ~300bp, despite seeing larger fragments (1kb) on the bioanalyzer.

To make better use of the 600-cycle v3, we want to try a size selection for Nextera. Has anyone attempted this? What did you try, and how successful were you? We were thinking of running the fragments on a gel either post-Zymo cleanup and isolating >600bp followed by PCR, or run the final library on a gel and isolating anything>600bp. Please share your experiences!

[kwaraska]

Generally speaking, nothing larger than about 700 bases will cluster.

I would recommend against a size selection because since the transposase is not completely random, you could be cutting out a huge chunk of your DNA.

I have heard that if you cut down on the DNA amount relative to the enzyme, you get smaller pieces-but that is anecdotal.

I would suggest calling Illumina-they have a tech support line just for library construction. They should be able to tell you how to decrease the insert size.

[sweetph3]

Yes, but we would like to see LARGER inserts. We'd like to see the 2x300 cycles give more data than reading a 300bp (or smaller) insert 100% in either direction.

Any other ideas or advice?

[lmw.bio]

Any luck?

Hello, I'm preparing to do some environmental sequencing on Illumina and also interested in sequencing large fragments with the 600cycle v3 kit. Just wondering what you ended up doing? I'm currently thinking of gel purifying all of the bands of interest before starting my library prep (NEB Ultra).

Thanks

[nucacidhunter]

Hello, I'm preparing to do some environmental sequencing on Illumina and also interested in sequencing large fragments with the 600cycle v3 kit. Just wondering what you ended up doing? I'm currently thinking of gel purifying all of the bands of interest before starting my library prep (NEB Ultra).

Thanks

If you are going to use sonication based fragmentation for NEB Ultra, I would suggest to use TruSeq Nano shearing and bead-based size selection for 550bp workflow and then use size-selected DNA with NEB kit. You might need to do minor adjustment to beads volumes depending on supplier.

[lmw.bio]

no shearing

Thank you for your reply nucacidhunter. I do not plan to do any shearing of my amplicons, instead was thinking of doing a size selection before the adapter and index ligation with the NEB Ultra kit. Amplicons range in size from 450 - 700bp but I am also amplifying some unwanted bands, mostly of much higher molecular weight. Would you still recommend using a bead based selection or gel purification?

Thanks!

[nucacidhunter]

Hi Imw.bio

Without seeing your amplicons profile, I would consider following:

1- Is the HMW material input DNA? If you quantified your input it will be easy to work out this by looking at concentration.

2- Multiplexing level which will affect output for each amplicon. Do you need every read for downstream analysis or can you tolerate some off target reads?

3- Number of samples: with low number of samples you will get clean cut from gels. If sample number is high then bead clean-up is more cost effective in expense of some off-target fragments going through sequencing (bead clean-up will include some smaller and larger fragments than your target size).

[lmw.bio]

Good call

Thanks again nucacidhunter! I appreciate the way you broke it down for me, that helped me see the big picture more clearly. I think I can afford the loss of some of my desired sequences in return for the gain in time and cost needed to do the gel purification. Thanks!

[jbarfield]

Sequencing bacterial DNA libraries on MiSeq using Nextera XT and V2 chemistryy

Hello,
I also have a question in relation to size selection for Nextera XT libraries and optimising the cluster density on the MiSeq flow cell.

I am finding the normalisation bead step to be producing inconsistent clustering. I was advised by several colleagues at a recent micro conference to skip this step and go back to the Nextera way ie. Qubit quantification post Ampure bead PCR clean-up and fragment size check using the BioAnalyser. Usually my libraries are an average fragment size of ~500bp which I use the formula 1 ng/uL = 3 nM to convert my Qubit values to nM conc (using Illumina's advice)

The next run, I plan to: dilute each sample to 4 nM and pool 5 uL of each. Dilute 80 uL pooled library with equal vols of 0.2N NaOH, take 5 uL of 2 nM denatured pooled libary and add 995 uL of prechilled Hyb buffer to produce a 10pM loading sample. Sample is denatured at 96 deg C, chilled on ice then loaded onto the MiSeq. Is it now the consensus to use equal vols of 0.1N NaOH instead of 0.2N NaOH?

Also, would it be worth doing a further bead size selection on each sample before dilution, pooling and loading my final sample? I have heard about SPRIselect beads that will capture a targeted region of specific fragment sizes (~500 bp) and exclude fragments above and below target sizes. Would I then need to do a final Qubit reading and adjust my nM conc before diluting to 4 nM? I would appreciate any advice on this please!

[nucacidhunter]

Hello,
I also have a question in relation to size selection for Nextera XT libraries and optimising the cluster density on the MiSeq flow cell.

I am finding the normalisation bead step to be producing inconsistent clustering. I was advised by several colleagues at a recent micro conference to skip this step and go back to the Nextera way ie. Qubit quantification post Ampure bead PCR clean-up and fragment size check using the BioAnalyser. Usually my libraries are an average fragment size of ~500bp which I use the formula 1 ng/uL = 3 nM to convert my Qubit values to nM conc (using Illumina's advice)

Bead normalisation is based on saturating the beads binding capacity with excess library fragments and it will be less efficient and inconsistent if PCR yield is less than 10-15 nM. In this case, qPCR based normalisation after cleaning PCR is more accurate. Average size for calculating molar concentration can be read from Agilent electropherogram by selecting the smallest fragment to 1kb range because fragments larger than 1kb rarely form clusters and are not quantified by most standard qPCR methods used for Illumina library quantification.

With SPRISelect beads a double SPRI size selection is necessary, but one does not need double SPRI unless downstream application requires more precise fragment length. A left side size selection is sufficient to remove fragments with insert shorter than sequencing cycles.