18s V4 region in Illumina MiSeq Amplicon

[18sSpace]

Hello everyone,
we are experiencing problems with our amplicon seqeuncing with a Illumina Miseq 250bp paired-end run.
We amplified our sediment samples with general eukaryotic primers for the V4 region of 18s (described in Bråte et al 2010) and send them in for library prep and sequencing. We now have for the second time a failed run and can't figure out what might be the problem. Our primers do not contain modifications, the sample was spiked with PhiX to increase sequence diversity and the samples are of high diversity to begin with. Cleanup was done with the chargeswitch kit. The quality of the amplicons was good (gel and nanodrop) and the concentrations (Qubit) were also high and more than sufficient. The library prep we ordered was done as a regular TruSeq adapter ligation.
Has anyone experienced problems with failed runs (mainly problems with R2 and indexing run) with a similar setup?
Thank you for your input,

[microgirl123]

What kind of QC is your center doing after library prep? And what kind of cluster densities are you looking at for your runs?

It doesn't sound like it's a problem with your amplicons but rather a problem with either your library prep or the run itself.

[18sSpace]

We just talked to them and they said about the QC that they run Bioanalyzer and Realtime PCR to determine amount, quality and adapter ligation, so the library prep works. But what they found out is that the ligation seems to be directed (FWDprimer always with P5) for some reason. We double checked if we have a modification that could explain that but we don't. The bigger problem seems to be that the PhiX spiking is failing (DNA for 40% in and 3 % sequences out) and therefor is the diversity too low to produce acceptable quality. As we are not the only ones using this PhiX spiking and it works for everyone else this is puzzeling. So to be honest, we are not quite sure what to make of this information and for us it didn't clearify anything.

[microgirl123]

What are your cluster densities like and are you using V2 or V3 chemistry? Is the center using the newest version of RTA software? How big is your amplicon (without adapters)?

It also sounds like maybe they are "under-quantifying" your sample so the phiX spike is not correct. However, if the newest version of RTA is being used, you should be getting good data from a 5% phiX spike.

[18sSpace]

I am back. Sorry, it took us a while until we got the answers from the sequencing center on the cluster densities. Thank you for your suggestions until now.
So, cluster densities are at 1071 +/-42 K/mm2. For this run they used the 2x250 kit, but a run before that (that seemed to have a similar problem) was the 2x300bp kit. I guess they are using the latest software version, but I am not sure about that. Our amplicon is 480 bp long without adapters.

[microgirl123]

Hmm... your cluster densities are a little high for a low-diversity run, but not so high that I can see your whole run failing. Do you have access to the run metrics on BaseSpace so that we can have a look at them? Also, have you contacted Illumina tech support? If you give them as much information as possible, they can look at your run and try to help you figure things out.

[cement_head]

I am back. Sorry, it took us a while until we got the answers from the sequencing center on the cluster densities. Thank you for your suggestions until now.
So, cluster densities are at 1071 +/-42 K/mm2. For this run they used the 2x250 kit, but a run before that (that seemed to have a similar problem) was the 2x300bp kit. I guess they are using the latest software version, but I am not sure about that. Our amplicon is 480 bp long without adapters.

Hello,

We've just completed a 16S amplicon run using a 10% PhiX spike-in. We followed the emp (Earth Microbiome Protocol) for the 16S, which is very similar to their 18S Protocol. You can find (working) details of the entire protocol, including the ISME paper (carefully read the Supplemental) here: http://www.earthmicrobiome.org/emp-s...protocols/18s/

I guess one thing that I liked about their protocol, before we even started it was the fact that there are no adaptor ligation steps - which I've always considered the Achilles heel of complex mol bol protocols. You should be able to design primers to incorporate the necessary sequences (by PCR) using your amplicon libraries as template. Actually, this is now becoming the preferred method (find attached reference Berry 2011)

- CH