GeCKOv2 CRISPR Library Deep Sequencing

[Daniele Simoneschi]

Hy everyone,

I have recently obtained the 2-vector system Human GeCKOv2 CRISPR Library from Addgene and followed the online protocol for amplification. As recommended, I would like to perform a deep sequencing analysis to validate library complexity before proceeding to the actual CRISPR screening experiments in the future. I have bought all 24 primers that you made available (F01-F12 and R1-R12) and would like to know if you could share the PCR protocol you used for validating the library complexity and making sure that each sgRNA has been maintained during bacterial transformation and Maxiprep.

For instance, what primers do you use? How much template (library) do you use per PCR rxn, and how many PCR rxns do you perform in total? What are the PCR parameters (Tm, extension, polymerase, final volume etc)? More generally, I would like to know how to proceed in order to prepare the library for NGS.

I would like to thank you in advance for your help with this matter. Looking forward to hearing from you soon.

Best,

Daniele

[Neville Sanjana]

Hi Daniele,

For validating the representation of the plasmid library, you can use any high-fidelity polymerase (recently, I've been liking NEB Q5). I basically follow the NEB manual for the PCR conditions.... except, you want to do the minimum number of cycles that allows you to see a band (i.e. prevent overamplification). Usually 15-20 cycles should be sufficient. For PCR from plasmid, just use a few ng of the plasmid as a template (1-10ng is fine) and an annealing temperature of 60C.

If this is the only sample you will be sequencing (Illumina), you will need to add stagger diversity.... for the F primer mix together F01 through F12. For the reverse primer, you can just use R01. This will create a diverse set of amplicons (due to the F primer stagger) that can be sequenced on Illumina. 

Hope that helps,

Neville

--
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 post to this group, send email to cri...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Daniele Simoneschi]

Hi Neville,

thank you so much for your reply. Your answer was very helpful and much appreciated. Just one last question; to save time on one PCR, I bought the concat primers as ultramers from IDT (about 90bp, including Illumina adaptors, stagger, barcode and priming site). Do you have any special recommendations when amplifying with those? My aim is to accomplish one single PCR, purify on a gel, and send everything to sequencing.

Thanks again,

Daniele

[Neville Sanjana]

Hey Daniele,

No special recommendations.... the PCR from the plasmid template should work well (Ultramers are of course better than standard desalt oligos) and your plan to gel purify sounds good. Just make sure to avoid overamplification by doing a few parallel reactions with different cycle numbers and choose the lowest cycle number that produces a clean single band.

Good luck!

Neville

[Marina Badenes]

Hi Neville,
I have a question about this topic.
I want to perform a genome-wide genetic screen using the mouse GeCKOv2 CRISPR library. After transduction and selection with puromycin we will sort the population of interest and control (parental population) to perform deep sequencing. What is the minimal amount of cells needed to do the deep sequencing?

Thank you
Marina

[Neville Sanjana]

Hi Marina,

Thanks for your question.... This number depends on the coverage that you want. For example, if you want 300x coverage (300 cells per sgRNA) and your library has 100,000 sgRNAs (this is roughly the number of the GeCKOv2 A+B libraries), then you want to sort/extract gDNA from 3e2 * 1e5 = 3e7 cells (30 million cells). 

Hope that helps,

Neville

[Marina Badenes]

Yes, it helps. But it will be challenging to get that amout of cells. Probably we will have to perform several sortings.

Thank you!
Marina

[Marina Badenes]

Hi Neville,

Can you advise me what do you think is the best aproach to perform this experiment? My population of interest will be 1-2% from the total number of cells so I won't have enough cells in one sort for sure. What do you recommend, sort the transfected cells, expand and sort them and so on, or expand several times the transfected cells and perform several sorts of them.

 I hope you understand my question

Thank you
Marina

[Daniele Simoneschi]

Hi Neville,

thank you very much for your continuous help. I am now ready to send samples to Next-Gen Sequencing for the last step of my screening. However, I was wondering if you could comment on the number of samples that you would multiplex on one single run of Hi-Seq. I know that using the barcodes we can reach 144 different combinations, but I am afraid that the more samples you multiplex the less coverage you obtain and, thus, the more uncertainty you have when selecting hits. For instance, would it be wise to run 20 samples on Hi-Seq; would this give enough sequencing coverage to each sample? What is the maximum number of samples that you have multiplexed?

Last, but not least, have you ever performed the final sequencing step of the screening on a Mi-Seq instead of a Hi-Seq? I'm asking because of pricing reasons.

Thank you so much for your time. Hope everything is well.

Daniele

[Neville Sanjana]

Hi Daniele,

No problem…. happy to help! Yes, we often do a MiSeq (use the v3 kit) first, which can give a good idea of whether a screen worked and/or top hits. Even with a few million reads per sample, you can have enough data to get going (and then later decide if you need higher read depth with HiSeq). 

With number of samples, it is up to you. Basically, take the total number of reads (20-30M for MiSeq, 100-200M for HiSeq) and divide by the number of samples to estimate per sample reads. For GeCKO-sized libraries, we often multiplex ~10 samples or so per run.

Hope that helps,

Neville

--

[Daniele Simoneschi]

Thanks Neville,

you are always so detailed in your answers, and I really appreciate all your help. I will try to run 5-10 samples on a MiSeq to see how it goes, and then decide whether I need HiSeq to get more sequencing depth.

I was wondering, have you ever frozen your cells after a large-scale infection and puromycin selection in order to have aliquots for later screenings. What if I apply the library, select them with puromycin in order to enrich for the KO population, and make aliquots to store at -80 for future studies by other people in the lab?

It's just a thought, any piece of advice is highly appreciated.

Thanks again,

Daniele

[Neville Sanjana]

Hi Daniele,

While we were working on our in vivo screen (Chen*, Sanjana* et al., 2015), we did try freezing (cancer) cells that had a library infection and, to our surprise, the representation seemed well maintained after the thaw. I do recommend that you sequence before/after just to validate that this performs similarly in your hands too.

Best wishes,

Neville