GBS and restriction enzyme choice

[kevpar]

I'm in the midst of putting together a sequencing project bases around GBS. The manager of my core facility has some concerns about the GBS protocol in terms of the choice of restriction enzyme and the tradeoff with sequencing depth. This is fine, and I'm happy with fewer cut sites since it will still provide thousands of markers.

However he has some different concerns in that he thinks that for GBS a frequent cutter enzyme is needed that generates tags around 500bp long. This is because the GBS protocol does not shear the cut fragments and he thinks that if the fragments are longer the cluster formation on the flow-cell will fail and sequencing yield will be very low.

Would the PCR step for the GBS protocol negate this potential issue?
Or is this even an issue to worry about with GBS?

[rosatoc]

Hi kevpar

Our core facility is now beginning to develop a GBS service. What have you learned over the past few years which would help us from making 'learned' mistakes?

Thanks very much

[Laval]

Hello Kevpar,
In GBS, if you use frequent cutter it will generate large number of small fragment and most of the sequences are too small to map correctly. But surly it gives you very wide coverage as frequent cutter has much more sites in genome than rare cutter. In other hand, rare cutter increase depth which is required for high quality SNPs but that also produce much larger fragment which are problematic for PCR. This also resulted in uneven distribution of SNPs on genome. Therefor people started using penta cutter like ApeK1 which is a mid way of frequent cutter and rare cutter. And it gives very uniform distribution of SNP because of partial methylation sensitivity of ApeK1 enzyme. Recently, a study using ApeK1 with additional selective bases in primer demonstrated SNP identification with high depth and uniform coverage. You should read the paper of improved GBS (iGBS) before starting with your GBS project- http://www.plosone.org/article/info%...l.pone.0054603

[JackieBadger]

A little off topic...but i saw a presentation recently that showed ~ 50% of the sequences for each sample are thrown away as various forms of garbage. So keep that in mind.

[rosatoc]

Hi
What program(s) do you like for in silico restriction digest of reference sequences?

[Mr. Pértille]

Hello everybody

Hello everybody, my name is Fábio
I'm hope to be useful to answer all the questions as possible.
But I think that i will make more asks than answers

Thanks

[Mr. Pértille]

I make an essay with Pst1 and some large framents are appearing (recommended is between 100 and 400bp-DOI: 10.1371/journal.pone.0054603) for chicken genome (Gallus gallus) digested
I am looking for a program for simulating cleavage entire genome (Gallus gallus), some online tools are available to small pieces as you can see at this web-site: http://molbiol-tools.ca/Restriction_endonuclease.htm

I found a program called "Automate RestrictionMapper", but it works from the command line. I have no tested it yet, because our server are in maintenance and I need this result urgently
Which genome you'll want to digest?
Tank you

Fábio

[Jayakumar S]

Hello Kevpar,
In GBS, if you use frequent cutter it will generate large number of small fragment and most of the sequences are too small to map correctly. But surly it gives you very wide coverage as frequent cutter has much more sites in genome than rare cutter. In other hand, rare cutter increase depth which is required for high quality SNPs but that also produce much larger fragment which are problematic for PCR. This also resulted in uneven distribution of SNPs on genome. Therefor people started using penta cutter like ApeK1 which is a mid way of frequent cutter and rare cutter. And it gives very uniform distribution of SNP because of partial methylation sensitivity of ApeK1 enzyme. Recently, a study using ApeK1 with additional selective bases in primer demonstrated SNP identification with high depth and uniform coverage. You should read the paper of improved GBS (iGBS) before starting with your GBS project- http://www.plosone.org/article/info%...l.pone.0054603

In GBS, how the rare cutter increase depth which is required for high quality SNPs?.

Thanks,
Jayakumar S

[SNPsaurus]

GBS was designed for low coverage sequencing of samples that have a good reference so that missing data can be imputed. The frequent cutter might produce 10M fragments, and only a random 1M of them will be sequenced in each sample. You could increase from 1M reads to 100M reads and get 10X read depth of each fragment, but most people would rather generate fewer fragments, aiming for 100,000 fragments that can be sequenced to 10X read depth with 1M reads. The best way to do this is to use a less frequent cutter.

[nucacidhunter]

In GBS protocol number and size of fragments in a library is determined by choice of restriction enzyme and PCR (limiting extension time). Using a rare cutter enzyme will result in less and larger fragments thereby reducing fragment number. Libraries prepared with rare cutter in comparison with frequent cutter when multiplexed and sequenced in same number of lanes will have higher depth for reads. Using a rare cutter such as enzymes with 8 bp recognition site used in RADseq may not produce a robust GBS library. That might be one the reasons why they have chosen a 5 cutter for GBS.

[Jayakumar S]

In GBS protocol number and size of fragments in a library is determined by choice of restriction enzyme and PCR (limiting extension time). Using a rare cutter enzyme will result in less and larger fragments thereby reducing fragment number. Libraries prepared with rare cutter in comparison with frequent cutter when multiplexed and sequenced in same number of lanes will have higher depth for reads. Using a rare cutter such as enzymes with 8 bp recognition site used in RADseq may not produce a robust GBS library. That might be one the reasons why they have chosen a 5 cutter for GBS.

Hi,
Thank you for the clarification.

Jayakumar S