High-weight bioanalyzer fragments after PCR

[atcghelix]

Hi Everyone,

We've been doing a dual-SPRI size selection prior to library prep to try to restrict our fragments to 200-500bp (hoping to sequence PE 100bp on a HiSeq 3000). After this size selection, our fragments typically look like this:



After library prep (using a Kapa LTP library prep kit) our bioanalyzer results typically look something like this:



These are from a 50uL PCR reaction in which we're using 2.5uL of each primer (i5 and i7) at 5uM. We're using 6 cycles of PCR. The input template DNA is 15uL at anywhere from 15 to 50 ng/uL.

Questions:

1. Anyone care to speculate what that fat tail of junk is starting at about 600bp? My instinct is that these are fragments where the library fragments are priming each other (overamplification symptoms), but I also think I'm not doing too many cycles and that I probably have enough primer?
2. Anyone have any idea if these are suitably sized for a HiSeq 3000/4000?

Thanks very much!

[jwfoley]

The tail is consistent with overamplification, especially if this is after dual SPRI, which should have removed it. But it's hard to answer whether you're doing the right number of cycles, as I haven't used your kit and you didn't say how you quantified your starting template.

I'd be nervous about sequencing it. If it's overamplified, I've had bad results from overamplified libraries (low % usable reads). If it's actually just good library products that happen to be too large, the limit for bridge amplification is somewhere around 1000 bp so there's a lot of junk in there that won't cluster.

If this is a precious library and you can't just start over, I guess you could try redoing the high-MW SPRI exclusion to see if you can get rid of the long tail. My guess is you can't because it's not actually high-MW, just ssDNA that migrates slowly in the Bioanalyzer due to secondary structure.

[atcghelix]

Thanks for the feedback. Starting template was quantified using a quant-it dsDNA assay kit (broad range) on a PerkinElmer Victor plate reader (comparable to qubit).

The sample is not precious--we're just trying to dial in our protocols at the moment to be compatible with the HiSeq 3000. So I can experiment some more in the next week. Would you try less input DNA, more primers, or fewer cycles?

[nucacidhunter]

I also think I'm not doing too many cycles and that I probably have enough primer?

To adjust primer amount you can run PCR reaction before clean up and look for small fragments in the same size as primers used. To identify primer peaks a no template reaction can be set up for comparison side by side.

[atcghelix]

To adjust primer amount you can run PCR reaction before clean up and look for small fragments in the same size as primers used. To identify primer peaks a no template reaction can be set up for comparison side by side.

Ah that makes sense. So you want to see some evidence of primer peaks after the PCR is finished?

[nucacidhunter]

That is easy way to optimise primer concentration in a reaction. One also should be aware of primer concentration range recommended by polymerase supplier.