Targeted sequencing of ~40Mb genome

[JoeChris38]

I am looking to do targeted sequencing on approximately ~40Mb of mouse genome (~2% of the genome) from low input DNA (~100pg). Equivalent to sequencing the equivalent of 2 exomes worth of DNA from 20 cells.

My current approach has been to design large primer pools to PCR amplify ~40Mb from ~100pg DNA input thus capturing and amplifying the required 40Mb in one step. I have had varying levels of success with this approach but I am now looking for alternative methods.

Has anyone been doing a similar sort of protocol? Or maybe have any ideas on how I could reproducibly capture and amplify 2% of a mouse genome from low DNA input?

Thanks!

[qingchang]

Is it possible to do WGA with MDA and then use your primer pool towards library construction?

100pg should be enough to do WGA.

[JoeChris38]

I have thought about this option but I think my lack of reproducibility from the primer pool is as a result of the primer pool itself and not due to the low template input. Though I will be attempting what you suggest in the next couple of weeks.

I am also going to attempt WGA (probably with Qiagen's REPLI-g kit i.e. MDA) but would like to try an alternative method of capturing the 2% of the genome I require (other than my current primer pool).

Any ideas on alternative methods of reproducible 2% genome capture would be much appreciated.

[qingchang]

Would REPLI-g Single Cell Kit followed by ION Exome custom design work for you?

[SNPsaurus]

You might try using the PCR primers to generate biotinylated amplicons, and then do a pulldown of indexed Nextera XT fragments from each sample, either one at a time or pooled. The pulldown efficiency will be low, but you could still multiplex plenty of samples in a lane.

That way, you do the PCRs with lots of input so better quality. You perform Nextera XT fragmentation on the samples and amplify them and pool to get sufficient quantities for hybridization. Might need to do 2 rounds of hybridization.

We did something similar with biotinylated BACs (http://www.genetics.org/content/189/3/767.short), but didn't have to deal with the low input. That might be too low for good Nextera XT libraries, but we use 1-5 ng on regular Nextera and get sequencing libraries out. Not sure if it will work but thought I'd toss out an idea!

[maoshigua]

targeted sequencing

Hi all,
Now it is not a problem to get the targeted sequencing data. But there are some general problems in dealing with them. For example, when you try to perform SNPs calling, which dataset you will use as the reference sequences (if there is no good genome sequence available)? the targeted sequences which you used to design the probes (CDS data) or the de novo assembled contigs? or other options? Does anyone have any suggestions in this case?

Many thanks in advance.

Mao