Methylation-sensitive restriction enzymes for GBS

[Pauline dlp]

Hi all!

I need help in selecting a restriction enzyme for GBS. Elshire's protocol for GBS calls for methylation-sensitive enzymes (They used ApeKI). He also underscores this in GBS workshops. I understand the use of methylation-sensitive REs is to filter out repetitive regions and to enrich genic regions, right?

Now here is where I am confused. I went out to search for literature for GBS done on insects (as I will be studying silkworm genomics), and found that some GBS studies use non-methylation-sensitive REs.

[1] Silva-Brandão et al in 2015, used PstI and MspI (double enzyme GBS) for oriental fruit moth. PstI and MspI are both NOT methylation sensitive.

[2] Dupuis et al in 2017 used PstI and MspI for GBS of the North American spruce budworm.

They both cited Poland 2012 for double digest GBS using PstI and MspI.


So does this mean methylation-sensitive REs are not required for GBS after all? What would be the consequences of choosing a methylation-insensitive vs. methylation-sensitive enzyme for GBS?

[nucacidhunter]

Methylation sensitive enzymes are more useful for plants as their genome contains large repetitive regions due to duplication events. PstI is methylation sensitive in plants as in plants methylation normally is in non-CpG context.

Choosing methylation insensitive RE can result in sequencing non-informative tags from repeat regions increasing the cost and has no other consequence. But this will depend on the genome and I do not think it would be a concern for non-plant species.

If you have access to Pippin instrument ddRAD would be a better choice.

[Pauline dlp]

Thanks for the reply, nucacidhunter! So it seems methylation sensitivity is more of an issue for plants. I tried doing in silico digests of my reference genome with several REs, and the methylation insensitive ones give a decent number of fragments for the target size (~100-200 bp).

We do have a PippinPrep, would bring up the possibility of using ddRAD. But i'm not sure I could do that since the project proposal stated GBS methods will be used.

[nucacidhunter]

GBS is a common name used for even non-restriction enzyme based methods. For instance, Illumina recently released a GBS kit that prepares libraries without RE o. ddRAD generally results in good coverage of tags as there is good control over tag numbers.