Nextera XT prep for metagenomes?

[DrMarineMicro]

Hello forum!

Cannot find the answer to this anywhere....

The Nextera XT kit info says it's for prepping "small genomes (bacteria, archaea, viruses), amplicons, and plasmids."

Has anyone used to it prepare a metagenome? The metagenome I'm interested in presumably contains fungal, bacterial, archaeal, and viral DNA. I am only able to recover small amounts of DNA (I would have to amplify it to use any other library prep kit, which I would prefer not to do given known biases in amplification).

So, it is possible to use the Nextera XT kit for prepare a library from a metagenome?

Thanks!!

[nickloman]

Yes, it is. At least, we've done it and it seemed to give comparable results to the same sample prepared with TruSeq.

[DrMarineMicro]

Great, that is excellent news, thanks so much for the reply! Are these metagenomes published yet? It would be great to have a reference for this grant proposal I'm working on....

[nickloman]

Not yet, although we have a manuscript in review. I'll update the thread when there is a citation.

[RCJK]

Nick, have you compared metagenomic DNA prepped with Nextera vs Nextera XT as well?

[bodington]

We're also interested in a comparison between the Nextera XT and Nextera DNA kits.

Personally I'd prefer to us the Nextera DNA kit, but our main problem is that many of our rare samples are in the 50ng-100ng range, leaving little room for error.

[lac302]

bumping this old thread.

i've been having a hard time getting decent cluster density out of metagenomic samples run through the nextera xt kit.

based on a fragment analyzer trace the distribution is quite large 400-2300bp. this was using 1ng input and .5x ampure beads.

which has a greater effect on size distribution dna input into tagementation or volume of ampure beads used for cleanup?

[nucacidhunter]

i've been having a hard time getting decent cluster density out of metagenomic samples run through the nextera xt kit.

It is probably an issue with correct quantification of library.

which has a greater effect on size distribution dna input into tagementation or volume of ampure beads used for cleanup?

That depends on how one looks at it. Using less DNA (high molecular weight) will result in smaller tagmented DNA reducing average size but the range will not change a lot unless one uses substantially less than recommended input amount. Bead is used to clean up PCR reaction and also to do a left side size selection (removing smaller fragments), which will dominate sequencing reads if not removed. Nextera libraries' size before clean-up is around 150-2000 bp. Cleaning with bead (0.5x) usually cuts off at 400-500 bp which then increases average size.

If I understand correctly your metagenomics library is prepared from a DNA sample of a community such as soil. I think that Nextera XT is an incorrect kit to prepare metagenomics library if the aim is to obtain good representation of organism’s genome. In such a community, conservatively, assuming there are 50 fungi and 100 bacterial species, there will be 50x50 Mb for fungal and 5x100 Mb for bacterial species totalling 3 Gb genome. Nextera XT is recommended for library prep mainly from one species with a small genome.

as lac302 mentioned, due to the wide fragment distribution (up to 2300bp), there's a lot of sequence one is actually throwing away. Fragments around 1kb have low clustering efficiency. This would cause a lot of bias in terms of species coverage in the metagenome, unless one can prove that the digestion pattern in terms of fragment length is the same for all species, so you'd miss out on big fragments equally for every species in the metagenome...

[lac302]

Originally Posted by nucacidhunter

It is probably an issue with correct quantification of library.



That depends on how one looks at it. Using less DNA (high molecular weight) will result in smaller tagmented DNA reducing average size but the range will not change a lot unless one uses substantially less than recommended input amount. Bead is used to clean up PCR reaction and also to do a left side size selection (removing smaller fragments), which will dominate sequencing reads if not removed. Nextera libraries' size before clean-up is around 150-2000 bp. Cleaning with bead (0.5x) usually cuts off at 400-500 bp which then increases average size.

If I understand correctly your metagenomics library is prepared from a DNA sample of a community such as soil. I think that Nextera XT is an incorrect kit to prepare metagenomics library if the aim is to obtain good representation of organism’s genome. In such a community, conservatively, assuming there are 50 fungi and 100 bacterial species, there will be 50x50 Mb for fungal and 5x100 Mb for bacterial species totalling 3 Gb genome. Nextera XT is recommended for library prep mainly from one species with a small genome.

I trust the Qubit quantitation here. I'm not using the normalization beads in this experiment since it is only one samle/library.

What preps would you recommend? Illumina tech support recommends doing a second AMPure bead cleanup to remove the right shoulder of the fragment distribution and then reanalyzing on bioanalyzer/fragment analyzer. If i have to do that I might as well make a few more libraries with varying input quantity to see the effect....with all this the size bias does concern me since we are looking at full length genes post assembly.

[nucacidhunter]

qPCR is the gold standard for quantification of Illumina libraries to obtain consistent read number from every run. With Nextera libraries a partial size selection occurs during clustering. The largest insert that I have sequenced has been 960 bp long even though there were fragments up to 2 kb in input library. I do not see much benefit in doing a double-SPRI size selection if one uses qPCR for quantification. If qPCR condition is set to amplify up to 1 kb and only that region from Bioanalyser is used for sizing (0.1-1 kb), then qPCR would provided more accurate results. I see that there would be a benefit for double-SPRI if one uses Qubit to quantitate mass because converting mass (which would be in effective clustering range) to molar concentration would give relatively good results.