ion torrent proton QC problem

[zcm2403]

Hi,everyone!
I had ran a p1 chip yesterday.After running completed i found that the on target reads and uniformity was very low,just as shown in the picture.Another problem is that my designed amplicon length is about 300 bp,however the results reads mean length is only 150 bp.There is something wrong in my experiments.could someone give me some suggestions?
THANKS!

[Xray1]

I have no Proton experience myself, but max read length of Proton is 200bp, isn't it?
What type of application is this, Ampliseq or something else?

[zcm2403]

thanks for your apply.My experiment is Ampliseq.And just as shown in the picture,the max length is about 350bp.I think the cause is that the long reads is too less.

[Xray1]

What panel? As far as I can tell, Ampliseq research panels (those designed by Ion or community panels) are usually 125-275bp amplicon range. These should work on Proton.

Since Pipeline Version 4, the Ampliseq designer allows you to use long amplicons with 125-375bp amplicons. These panels will only work on PGM or S5-520/530 with HiQ-Chemistry. These panels are nice, as you usually get better design coverage. But they have these limitations (also not usable with equalizer kit and therefore probably not usable with IonChef Ampliseq library prep). Ion/Thermo likes to forget about these limitations in their documentation, like no word on this in the Ampliseq Kit documentation. But in the Ampliseq designer, it will tell you what machine/chip you can use the panel.

[zcm2403]

THANKS!I had used the HiQ kit with 375bp amplicons on the proton before.The results was good.I do not understand why the on target reads and the uniformity was so bad in this panel?I mixed some short length pannel with long length panel.Is it the keypoint?

[Xray1]

When I get uniformity near 50%, it is usually just that one of two primer pools didn't work. You can check whether amplicons with low coverage are all from the same pool. If so, your ampliseq library prep just didn't work.

Mixing libraries generated with long and short amplicon panels in the templating step might add to the problem. However, most "long amplicon" panels also have shorter amplicons and still work just great on the PGM. I am currently running a long amplicon panel with the shortest amplicon somewhere around 130pb and it shows great results. However, for Chef/PGM (and I think also the OT2) templating settings are different for 200bp and 400bp runs. For Proton you probably always use the same templating and that might cause more proplems when you add more short amplicons. Make sure a certain long amplicon panel works fine on the Proton before mixing with other panels.

You will have to try around a bit I guess, since the long panels are not officially supported. Interesting to read that it can work. Getting longer reads on Proton was one of the things people really hoped for (besides new P-chips.....), so the fact that Ion never got anything going (not even on the S5-540) indicated to me that it just didn't work. I guess most people therefore just avoid even trying this......Do you adjust any of your protocols (amount of library you put in or whatever) to get the long amplicons running?

[zcm2403]

Thanks for your suggestion.Before running on the proton i did a preliminary experiments on the PGM.And the uniformity was about 85%.So the results was good.Longer reads panel works well on the proton in our lab.And except the HiQ kit we did all the experiments following the protocals.

[Isabella97]

Thanks. I was looking for this answer for a long time.