I have been using Nextera XT library prep kit to make libraries with my amplicons. My amplicons are 1 kb to 8.6 kb long. I ran gel to check the DNA before and after Ampure XP bead clean up, before library prep, and they look fine (Figures 20190828 L13 K2G A-D and 20190828 L13 purified K2G A-D). I quantify the purified DNA with Picogreen and dilute to 0.2 ng/ul. In my library prep I half the volume of everything. The tapestation shows that the average fragment size is ~300 bp. After sequencing on MiSeq (2x 150 cycles), the reads have a lot of overrepresented sequences and adapter sequences.
I have done a few things to tackle this issue. First is to strictly control the 55 C incubation to 5 min exactly. It didn't help much. Then I look into the DNA quantity. I actually diluted the input DNA to 0.2, 0.4, 0.6 and 0.8 ng/ul and added 2.5 ul of DNA (equivalent to 0.5, 1, 1.5 and 2 ng) and 2.5 ul of ATM. This was done on the same plate 3 columns for each DNA input. The fragment size indeed improved a little when I put in more DNA but still smaller than expected (Figures size comparison and tapestation). The reads quality was still not great.
I also quantify my diluted DNA with Picogreen. I found that at lower concentration, Picogreen seems to overestimate (gives higher readings). For example, I had the DNA of 4 ng/ul (measured by Picogreen), I diluted 20x. The diluted DNA should be 0.2 ng/ul. When I measured the diluted DNA with Picogreen again, it gave the value of 0.3-0.6 ng/ul.
I gave my amplicons to our collaborator for library prep and their result was great. It ruled out the possibility that the amplicon sequence causing the problem (for example the Tn tented to cut more on some sequences)
I don't know what else I can do to improve my library prep.
After trimming the adapter sequence and low quality bases and filtering the reads < 50 bps, I assembled the reads. The assembling is kind of OK but the coverage is very uneven (Figure uneven coverage). I am not sure if this is caused by the short fragments in library prep. I don't find the uneven coverage improved with longer fragment size (400 bp vs 300 bp). May be it is not long enough? Instead, I found that usually in a plate, Row A/H samples have serious uneven coverage problem while the samples in the middle are a little better.
Any suggestion is highly appreciated. Thanks!
You may want to size select your amplicons on a low melting gel before PCR or reduce the ratio of volume of Ampure XP beads in your cleanup prep. You can test the latter with a size marker first to find the right volume of Ampure beads to use.
David John Kowbel Thank you for your help!
I have been using 0.6x beads to purify both my PCR products and the amplified libraries. I found that there are two issues in my samples: too small fragments as well as too many overrepresentated sequences. I am investigating the properties of the overrepresentated sequences and hopefully find a hint on what I can do next.
I think it's time to wrap up this question and hopefully it can be helpful for others. I finally got the answer: it's the thermocycler! My collaborator and mentor Professor Sarah Palmer suggested to try on other thermocycler after hundreds of different troubleshooting experiments. First time I actually performed the library prep in her lab using an Eppendorf thermocyler (forgot the model) and it worked in my hands! Then I contacted Bio-Rad trying to organise the technician to do calibration for our T100. I was told by the company that T100 is a very basic model not suitable for library prep. I should use C1000. Our lab manager managed to find a Bio-Rad C1000 for me. I ran my library prep on it and it worked!!! Phew~~~ never thought that it was the thermocycler. It has been such a long time and frustrated process but at least it is solved. Hope this lesson is helpful for anyone experiencing similar problem.
- Badges
- Science method