% PhiX for Illumina sequencing protocols and SAV interpretation?

[GSeq94]

Hello,

I am a bit confused on how to calculate the volume of PhiX to add to the sample prior to sequencing on the Illumina platform. Based on the Illumina protocols for the MiSeq for example, If I am loading a total volume of 600 ul of which 300 ul of denatured pool and 300 ul of HT1, am I supposed to calculate the volume of PhiX on the total volume of 600 ul or based on the volume of sample I am using, in this case 300 ul? Based on the protocols, Illumina says to add 6 ul of PhiX in a 600 ul of final solution to achieve a PhiX spike of 1%. However, when I did this, on the SAV files, the % of aligned PhiX is always around the double of the amount. I am a bit confused because I am dealing with incredibly low diversity libraries that will yield to low reads passing filters, and would like to avoid adding a 30% PhiX (thinking of adding only 15%) because the PhiX will take up even more reads reducing again the amount of sample reads. I would really appreciate if anyone could help me understand.

Thanks and happy sequencing to all of you!

[BasesBreaker]

Hey there GSeq94

In my opinion, this is a pretty common problem. The actual PhiX spike-in percentage, as shown on the SAV files, tends to differ from the expected PhiX percentage.

Make sure when you're adding PhiX to your samples, the percentage of PhiX is calculated based on the total volume of the denatured pool you are loading onto the MiSeq, not the final volume after mixing with HT1. For example, if you're adding 6 µl of PhiX to a 300 µl denatured pool, that's 2% (6 µl / 300 µl * 100%). However, after adding HT1, the volume will be 600 µl but the PhiX concentration will still be considered 2% of the denatured pool.

I always experiment with different PhiX spike-in percentages to find a better balance for my specific libraries. You should probably start with 15% as you suggested. But this always seems to change depending on the samples and library type.

[MicaMarg]

Hello GSeq94 !

In my opinion, there's no change when you add the PhiX except that you'll have to do it after the denaturation on the MiSeq systems (I usually use aliquots of already prepared PhiX library at 20pM that I use in a single month). But the percentage has to be calculated for a targeted final concentration (not only the volumes if the PhiX concentration differ from your library concentration) ; so you'll have to change the dilution volumes dependantly.

At the end, I've usually seen 0,5-1% less PhiX than targeted. When 2% is targeted, I'll have 1,5-1,8% of Phix ; when 15% is targeted, I'll have ~14%.

We here are adding 15% PhiX for 16s metagenomics at 2x300bp sequencing (=really low diversity), targeting 800-1000K/mm2 clustering ; it works, but we here are pushing the system at it's limits...



Tell me if anything work or not for your needs... Best regards

[GSeq94]

Dear BasesBreaker and MicaMarg thanks for your reply. May I ask the exact reason why I would need to add the PhiX before adding the HT1? I think this is the correct way as well, but if we red the Illumina guide on how to use the PhiX (screenshot attached) you can see that a 1% spike corresponds to 6 ul of PhiX on a volume of 600 ul of denatured and diluted pool (594 ul + 6 ul PhiX).

For Low diversity libs we also tend to do 15-18% spikes for 2x300 and we get around 800 K/mm2.

Thanks

Giulia

[MicaMarg]

You can add the Phix when you wants to, but as it's logically the less representated library it's more logical to add it at the end (the PhiX added volume will be low...). You can even mix your libraries before the denaturation, with fresh thawed PhiX, and denature the whole mixture then, as long as your calculations are right (based on C1V1=C2V2), and your dilutions too, but it's more complicated.

Here in the MiSeq Denature and Dilute Guide, it's not written on the screenshot but the volumes are given thinking that the PhiX's concentration is the same than the libraries have. But if your libraries are more concentrated than the PhiX, you'll have to add more of this one. Mathematically