From john.mcewan agresearch.co.nz Sat May 12 23:01:23 2012
From: "McEwan, John" <john.mcewanagresearch.co.nz> (by way of Jill
Maddox <jillian.maddoxalumni.unimelb.edu.au>)
Subject: RE: Minutes from Sheep Genomics Consortium Call #100
To: Multiple Recipients of <sheepmodelsanimalgenome.org>
Date: Sat, 12 May 2012 23:01:23 -0500
Dear all
Thanks for the valuable discussion on this
thread, that is exactly why I raised it at the
ISGC meeting because I know many cannot attend, but many do read the minutes!
Here are my views (further comments appreciated):
¡¤ GBS and genome sequencing are the
future as Xun states. They are the future now
for: species with high sequence diversity and/or
likely low numbers of individuals needing to be
genotyped (read fish, poorly studied mammals and
plants). This will mean DNA sampling and
extraction methods will need to be modified at
least in the short term. As a consequence we are
very interested in GBS in a variety of species.
See for example
<https://pag.confex.com/pag/xx/webprogram/Paper2574.html>https://pag.confex.c
om/pag/xx/webprogram/Paper2574.html
¡¤ For species with existing genome
assemblies, and moderate volume of individuals
per annum, chips have and are likely to be priced
below GBS for at least the next 3-4 years based
on technology ¡°on hand¡±. Where GBS/RAD
sequencing is most competitive is in the ¡°low
density¡± region 10-20K SNPs. It *currently*
struggles price wise where you want say 700K SNPs
and a defined content. This leads to say rice
researchers sequencing accessions, then using
high density chips on additional plants, and then
using GBS/RAD sequencing for low density genotyping on many more individuals.
¡¤ It is recognized that SNP chips are
not as flexible as GBS (start up costs, can use
same protocol across closely related strains and
closely related species) and avoids many issues
about SNP selection and ascertainment bias.
¡¤ It is recognized that chips are less
fussy about DNA quality and are easier to process
regards volume constraints (i.e. you can run say
24 animals and get results within a week). This
makes subsequent commercial use *in animals* vastly simpler.
¡¤ Many ovine users around the world have
invested significantly in 50K genotyping and want
to transparently upgrade by imputation to higher
densities, this means it is a significant
advantage to have the 50K SNPs on any higher
density platform at least initially.
¡¤ We have a need now, not in 5 years and at a specific price point.
¡¤ So this gets back to Xun¡¯s very
generous offer. If Xun can do GBS for >500K
SNPs/individual sheep (i.e. >8% of the genome)
using some form of reduced representational
sequencing method for US$50-$75 now we should
seriously investigate it before committing to a
HD chip. At that cost we could also GBS a
subsample, perhaps 5-10%, of sheep already done
on 50K chips (estimated as ~60-70,000 total to
date) to impute up the balance as well as proceed
with GBS going forward on the new animals
mentioned below. This leaves the LOF SNPs
dangling but that is a tradeoff that can be made
and there are alternatives to placing these on an HD chip.
Cheers
John McEwan
.From: Stephen Moore [mailto:s.moore3uq.edu.au]
The situation with arrays is changing rapidly as
well. In addition to a potential significant
reduction in price the ability to economically
design and manufacture high density custom arrays
looks like being a reality in the near future.
Unless sequencing drops below $30.00 I think the
arrays still have a few years life in them.
.From: Claire Wade
<<mailto:claire.wade@sydney.edu.au>claire.wadesydney.edu.au>
Hi all,
while I concur with Xun that genotyping by
sequencing offers many advantages over the use of
arrays, one of the big problems with it that we
have been struggling with is the requirement for
high quality DNA. Buccal samples will be
unsuitable due to bacterial contamination, blood
samples need special care. Buccal samples are
also unsuitable for RAD due to their alteration
of DNA fragment sizes as a result of
degradation. Sequencing really is a garbage in
garbage out situation with (Illumina) arrays
being quite robust to poor DNA quality. The
ongoing availability of arrays does remain an issue of concern.
Claire
.From:ÐìѶ [<mailto:xuxun@genomics.cn>mailto:xuxungenomics.cn]
.Subject: Re: Minutes from Sheep Genomics Consortium Call #100
Hi James and all,
Sorry that I did not able to attend the consortium meeting for a period.
I notice in the argument for genotyping platform.
I think the current progress of sequencing
technology is with many advantages beyond array
based genotyping. GBS and RAD will be a good
choice for molecular breeding, with much cheaper
cost and more efficiency workflow. If you still
feel that those markers are not meet your needs,
another choice will be target region capture and
sequencing, which will specifically target those
700k SNP sites. However, I still believe GBS and
RAD will be the best choice and practice for
molecular breeding. BGI would like to help on
those. How about the following collaboration:
(1) BGI will design the capture array, target 700K SNP sites.
(2) BGI sequencing 7000 samples for the
consortium for these 700k SNP sites by the price
of 300USD (the same price as illumine offered for only Chips)
(3)BGI will do BGS for all these samples by our own cost.
(4)The consortium and BGI will work together to
compare if GBS can replace those high density
genotyping in breeding and other practice.
Best,
Xun
On 12-5-8 ÉÏÎç10:47,
"<mailto:James.Kijas@csiro.au>James.Kijascsiro.au" wrote:
Minutes from ISGC Conference Call #100
2. Whole Genome Sequencing Experiment: HD Chip
John promoted discussion about the logic of
constructing an HD chip in light of the rapidly
changing sequencing technology and genotype by
sequencing (GBS) methods. He put the view that in
light of the large effective population size for
most sheep populations, that it would difficult
to see how existing methods could effectively be
used to sequence sufficient animals (tens of
thousands). To achieve compatibility with
existing SNP50 datasets, some combination of GBS
mixed with imputation would be required, which
again may be difficult given the high Ne in
sheep. Hans noted that imputation from SNP50 up
to whole genome sequence would benefit from an
intermediate such as an HD chip, while noting
that the GBS technology is advancing very
quickly. Gwenola indicated that an HD chip would
be useful, and no one across the group voiced a
strong opinion against proceeding with an HD chip.
James provided feedback about the communities
likely uptake of an HD chip. Illumina set up a
web based collection point for registrations of
interest to purchase an HD chip at three price
points. The result were that at $120 USD /
sample, interest was received to genotype 27,700
samples. Assuming $200 USD / samples, this
decreased to 13,800 samples and at $300 USD /
sample it went down to 6,800 samples. He then
characterised the strengths of the two providers
of HD genotyping. No decision was taken regarding
the choice of provider. John indicated that price
is his primary driver, above technical
perfection. The group will work towards a
deadline of December 2012 to have chips manufactured and distributed.
The team discussed design criteria for a chip
housing 700K SNP. John suggested as a starting
point inclusion of four sets of SNP:
1. SNP on the existing SNP50 Beadchip
2. Special interest SNP. This might include
GDF8/myostatin, monogenic disease SNP, PRNP SNP,
markers across the MHC etc. Jill is working to compile a likely list.
3. Minimum MAF SNP across breeds. These to be
identified primarily from the diversity panel of 75 re-sequenced genomes.
4. Exonic and Loss of Function (LOS) SNP.
Following SNP annotation against the available
gene models, SNP to be include that are predicted to ablate protein function.
The group decided that a 1 page design document
needs to be distributed, debated and finalised by
consensus. This seeks to ensure that everyone has
input and we construct a chip that serves the
communities best interests. John will write a draft and distribute.
----------
Attention: The information contained in this
message and/or attachments from AgResearch
Limited is intended only for the persons or
entities to which it is addressed and may contain
confidential and/or privileged material. Any
review, retransmission, dissemination or other
use of, or taking of any action in reliance upon,
this information by persons or entities other
than the intended recipients is prohibited by
AgResearch Limited. If you have received this
message in error, please notify the sender immediately.
----------
|
