Amplicon sequencing with 5% phiX spike

[microgirl123]

Has anyone tried low-diversity amplicon sequencing (16s) with the new Illumina recommendation of only a 5% spike of phiX? I have a couple of people who want to start prepping their libraries and are wondering if they need to artificially add diversity or not. Thanks!

[kwaraska]

I was told the 5% recommendation was only for MiSeq.

The HiSeq's (2000 and 2500) still need a higher percentage.

[microgirl123]

Sorry - forgot to specify that I'm running a MiSeq!

[kmcarr]

Sorry - forgot to specify that I'm running a MiSeq!

This recommendation only applies to MiSeqs running the latest version of MiSeq Control Software and RTA. Make sure your MiSeq is current.

[JackieBadger]

Depends on the complexity of your amplicon library.
From our own experiences, if you do not have a complex library you will need to add complex sequence to the 5' of your amplicon or use more phiX

[thomasblomquist]

I was familiar with the 5-10% for miseq. Do you need a higher percentage for hiseq? I was of the mind it should scale proportionally.

-Tom

[kmcarr]

Depends on the complexity of your amplicon library.
From our own experiences, if you do not have a complex library you will need to add complex sequence to the 5' of your amplicon or use more phiX

Jackie,

Have you tested this with the recent release of the MiSeq software which addresses issues with sequencing amplicon libraries?

[thomasblomquist]

Jackie,

Have you tested this with the recent release of the MiSeq software which addresses issues with sequencing amplicon libraries?

Is this the release I've heard of where the RTA won't kill the entire run if it can't focus on a particular step, or where/when, if possible, the RTA will define colonies/clusters later if cluster differentiation is better at later flows of nucleotides?

-Tom

[kmcarr]

Is this the release I've heard of where the RTA won't kill the entire run if it can't focus on a particular step, or where/when, if possible, the RTA will define colonies/clusters later if cluster differentiation is better at later flows of nucleotides?

-Tom

Not quite. Here are the Release Notes. Specifically it is the changes added in RTA v1.17.28 which make significant improvements to amplicon sequencing without need for low cluster densities, hard coded matrix and phasing values or very high (25-50%) PhiX. Recommendations for amplicon sequencing with this version of the software are standard cluster densities of 500-1200K/mm^2 and only 5% PhiX.

We have done a couple of runs thus far using the Caporaso & Knight primers. We had 16-18 M raw clusters, 13-15 M PF clusters, 93% ≥ Q30 on read 1, 87% ≥ Q30 on read 2 (PE 150 reads). We were shooting for 5% PhiX but ended up ~10%.

[thomasblomquist]

Not quite. Here are the Release Notes. Specifically it is the changes added in RTA v1.17.28 which make significant improvements to amplicon sequencing without need for low cluster densities, hard coded matrix and phasing values or very high (25-50%) PhiX. Recommendations for amplicon sequencing with this version of the software are standard cluster densities of 500-1200K/mm^2 and only 5% PhiX.

We have done a couple of runs thus far using the Caporaso & Knight primers. We had 16-18 M raw clusters, 13-15 M PF clusters, 93% ≥ Q30 on read 1, 87% ≥ Q30 on read 2 (PE 150 reads). We were shooting for 5% PhiX but ended up ~10%.

Thank you for the feedback. We ended up at around a 5-10% spike-in of PhiX to allow for appropriate cluster identification for our amplicon sequencing efforts.

Regards,

-Tom