Discrepancy with gel and tapestation

[Ariel Stanhill]

Dear Julia,

Thank you very much for your assistance with the GeCKOV2 CRISPR screen.

I have purchased the GeCKO v2 library for a FACS screen and after the amplification (steps 18-31 provided in your protocol) I have moved  to evaluate the library amplification by a 100x NGS sequencing (step 37)

After performing the PCR step, combining, purifying and running 2ug of  the sample on a 2% agarose gel (see attached gel picture) the sample of library A and library B look as expected (between 260-270bp). When I extract from the gel the ±270bp band  I performed a tapestation evaluation prior to the Qubit quantification of the  gel extracted band. The results indicate the DNA is of 431bp with minor amounts at ±260bp (see attached picture)

I have repeated this several times and get the same problem regardless of the index reverse primers I use. I even tried to use DNA beads for 

purification as not to use any denaturing reagents for the agarose melting but get the same results.

I have run once the samples and the NGS result indicated  their aren’t enough reads from these indexes.

I thought of maybe trying this on the original DNA of the library, to see if the amplification step I performed was problematic or at least to confirm the problem is not in the amplification step,  but with the amount of material I received from Addgene this will eliminate my ability to re-amplify the library if desired.

I would appreciate your assistance in this matter

Yours truly,

Ariel Stanhill

 

[Julia Joung]

Hi Ariel,

I vaguely remember seeing another post on this forum about the same problem, where the agarose gel results don't align with the Tapestation results, but can't remember quite what the reasoning was. I believe the solution was to move forward with sequencing because the agarose gel results look clean, and the Tapestation results indicate larger fragments for reasons other than your fragment is not the right size. Since you have repeated the PCR a few times and get the same result, I would suggest moving forward with NGS using the Qubit quantification and the fragment size you observed on the agarose gel.

Best,

Julia

--
You received this message because you are subscribed to the Google Groups "Genome Engineering using CRISPR/Cas Systems" group.
To unsubscribe from this group and stop receiving emails from it, send an email to crispr+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/crispr/31E66548-D675-47F3-8F96-9A3910640540%40gmail.com.

[Mithun Das]

Hi Ariel,

Perhaps spike-in a little bit of your library with other's sequecing run just to check whether you get expected amount of data. If you get expected amount of data compared to the amount of library you are spiking in you can go ahead with the final run. 

Between, how many cycles did you use in your first and second pcr reactions?

Cheers

Mithun

--

[Ariel Stanhill]

Hi Mithun

Thanks for your reply.

I am using the commercial GeCKO v2 library I purchased from Addgene.

I performed the pcr reaction exactly as described in step 33 of the nature protocol paper (22cycles). 

The NGS was performed as described in step 37.

I thought of trying to get another library from someone else and then see if the problem was in the library amplification or in the pcr. 

I did actually try to sequence it to see what I get, but there were very little reads from my specific barcode

Still can’t make sense of it 

Thanks

Ariel 

On 9 Nov 2020, at 15:43, Mithun Das <mithunda...@gmail.com> wrote:



<Lib PCR gel.jpeg>

<Tapestation.jpeg>

[Mithun Das]

Hi Ariel

This is exactly what I thought.

I doubt that the issue is library. I think there is something to do with the library amplification. If you try to quantify the library using PCR you will see that the library will not amplify at the expected level, although we should not expect that the amplified library will be as good as illumina standard library since we are using customized long primers, but it should not be 99 times low either. Perhaps try to reduce the pcr cycle number. Please keep in mind that different PCR master mix may have different efficiency. Depending on PCR master mix you uave to adjust the cycle number, but in general if you reduce the number of PCR cycle it should give you comparatively better yield.

Being said that I was sharing my experience based on two step pcr protocol. If you are using one step PCR protocol from nature paper this may vary.

Good luck

Cheers

Mithun 

[Yijun Tian]

Keyword: PCR heteroduplex. Very common for mix template amplification. PMID 11972349

To view this discussion on the web visit https://groups.google.com/d/msgid/crispr/CAC0WrT147_OaMgSMyA92-8xSnCPtuCprUtsfGWJxMM%2Bus1vzrQ%40mail.gmail.com.

[Xiao Lei]

Hi, Ariel and Julia,

I have a similar problem. I pooled my CRISPR PCR amplified libraries from different FACS sorted out cell populations together for both Miseq and Nextseq sequencing. While the Miseq pool looks very good (expected size around 276 bp), the Nextseq pool looks weird (main peak at around 350 bp). What surprises me is that these two pools are from a mixture of the same samples with just differences in ratios of individual samples in the mixture (I prepared the Nextseq pool more carefully to make sure each individual libraries are in equal amount nanograms, while for the Miseq I just mix equal volumes of individual libraries instead of controlling for equal nanograms). While the Miseq results are good, my Nextseq running has problems (I am not told the details of the problem but the sequencing facilities are doing troubleshooting for my Nextseq run these days).  From my experience, it seems that the problem is not causing by gel purifications (I heat the gel in the extraction buffer 55 Celcius degree for as long as 1 hour), but something weird occurred after the storage of these libraries (I prepare the mixture of Miseq and Nextseq at different times).

I attach my tape station results here and will update this post when I have more information.

Best,

Xiao

[Xiao Lei]

Hi, Yijun,

Thanks. But the problem is that we cut the gel at 270 bp (not the heteroduplex at 500 bp), but the tape-station still shows a 500 bp main species. Some people in this forum suspect it is due to sample denaturing during the gel extraction process, but someone in this forum also reported that beads purification (should keep samples in the native state) did not help in this case either. 

Best,

Xiao

[Xiao Lei]

Hi, All,

Here I attached my troubleshooting experience with the situation in which my CRISPR library's size is not consistent between gel extraction and tape-station and my sample failed Nextseq sequencing. I have a pooled CRISPR library (a mix of 10 individual FACS sorted and unsorted samples) that failed Nextseq sequencing. I was told that sequencing failed due to over clustering, the facility repeated the Nextseq sequencing with a much lower loading amount but failed again. To start troubleshooting, I run tape-station of each of my CRISPR libraries individually and noticed that the last sample (FACS2nd_P4) shows the wrong size in tape-station (the main size peak around 500 bp, although agarose gel showed the main portion at a correct size around 270 bp). I did a second round of gel extraction for the problematic sample, and surprisingly,  although I was so sure that I cut out the gel band at 270 bp, the tape station still showed the main peak is around 500 bp. It seems the tape-station results did not improve at all after the second round of gel extraction! However, when I mixed the second round purified sample with the other nine samples (50 ng each sample), the tape-station of the pooled library seems improved (main peak around 274 bp instead of 354 bp). I also tried to mix the 10 samples (each 50 ng, including the second round purified sample) together first and then did a gel extraction, this also improved the tape station results of the pool library (main peak around 281 bp instead of 354 bp). I sequenced the two preparations mentioned above in Miseq Nano,  both Miseq Nano sequencing worked but the clusters passing filter (CPF) were low (around 50-60%, a good CPF is >80%). The sequencing facility told me that my library was low diversity, and suggested to spike 30% PhiX with my sample in my Nextseq sequencing trial. What I finally did was that mixed the nine good samples (each with 50 ng) with half amount (25 ng) of the repurified problematic sample (FACS2nd_P4), and then do the Nextseq sequencing with 30% PhiX. This worked well finally and I had satisfactory results in the following data processing. 

Long story short, based on my experience, I would suggest that 1) check each library individually by tape-station, instead of just checking the pooled library. We may find one or two samples are problematic. 2) redo gel extraction of the problematic sample (this may not improve the tape-station results, but at least it likely helps in my case). 3) mix half amount or less of the re-pufried sample (or problematic sample) to the other good samples. 4) Increase PhiX to 30% in Nextseq sequencing (even you have a good library, 30% PhiX does not hurt). 

Best,

Xiao

[Valerie Prouzet]

Dear Ariel,

I have exactly the same problem: Nice Gecko PCR band at 270 bp on agarose gel after purification steps (gel extraction without warming, Ampure beads) but a TapeStation abnormal with major band at 500 bp. So I am interested to know if you could solve your problem since last november. I am afraid to make a NGS with my samples even if the agarose gel profiles are good....

Best regards,

Valérie