Dear Colleagues,
I am in the process of preparing a grant application to develop a
whole genome microarray (WGA) for Chlamydomonas and would like to get
feedback from everyone as soon as possible. I believe that such an
array would be an enormous asset for our research community, and the
more feedback I obtain the easier it will be to have the array made
and the more accessible it will be. Below I describe what is a WGA,
what it can be used for, and why we need one when we already have an
expression array. I also discuss cost and accessibility.
What is a WGA?
Whole genome arrays are made by fabricating oligos (usually
25-30mers) on a chip that cover as much of the genome as possible on
both strands. Space constraints can limit the coverage, and
repetitive and poorly hybridizing sequences are not included. In
practice there should be about one oligo within every 60 bp window of
genome sequence. There is no bias or a priori assumptions about gene
models or other annotation (although some bias could be deliberately
introduced in order to capture well-annotated exons).
How are they made?
Currently the state-of-the-art technology is controlled by
Affymetrix. They design a photolithographic mask that allows them to
simultaneously synthesize multiple arrays on a single wafer and then
subdivide the wafer into individual chips. The arrays are generally
better quality and can have a much higher feature density than glass
slide arrays. There is a substantial design and fabrication cost for
the mask, but once it is made it can be used repeatedly to make large
numbers of chips. Importantly, the fixed up-front costs can be
negotiated down or eliminated if a research community can demonstrate
interest in using the arrays, which is why I need your feedback. The
less money that goes into design and fabrication of the mask, the
more will be available to help subsidize the cost of the chips for
your experiments.
What can you do with a WGA?
Gene expression:
WGAs have a number of uses including quantitative gene expression
analysis. Compared to glass slides there are some distinct
advantages. 1) Quality and consistency of Affymetrix chips are
better than glass slides meaning that fewer replicates are necessary
to obtain a statistically meaningful result. 2) Typically each
transcript can hybridize to multiple features (oligos) on the chip.
This gives a huge boost in statistical power and reliability. Data
for one or more aberrantly behaving features can be thrown out and
you will still get accurate expression information from the remaining
oligos. 3) Instead of 10,000 predicted genes you get the whole
genome with no bias. 4) Small non-coding transcripts can be
detected.
For many types of experiments the glass slide arrays will be more
cost-effective. However, the additional applications listed below
cannot be done (or done nearly as well) using glass slide expression
arrays.
Annotation and gene prediction:
Those of you who have spent time annotating know that gene prediction
programs and ESTs have serious limitations. A WGA will give us a
much more complete and accurate picture of what's expressed and where
transcripts begin and end. Alternative splicing and non-coding
transcripts can also be detected. WGAs are not a panacea for
annotation since there is still potential for artifacts; but in other
organisms where they have been applied, WGA experiments have
identified a large number of novel transcripts many of which were
independently validated by other means. It will also be feasible to
interface WGA expression data with the genome browser.
DNA associated protein localization:
ChIP-Chip (Chromatin Immunoprecipitation followed by WGA
hybridization) experiments allow identification of all chromosomal
binding sites for a protein of interest. For transcription factors
it is especially powerful because ChIP-Chip data can be combined with
microarray expression data (e.g. expression profiles from a
transcription factor mutant versus wild type) to find the direct
targets of the transcription factor.
RNA binding proteins:
Similar to ChIP-Chip experiments except that all the RNAs associated
with an RNA binding protein are detected.
Mapping point mutations and polymorphism detection:
WGAs are very good at detecting polymorphisms between strains. If
S1D2 DNA were hybridized to a C. reinhardtii WGA and compared to
hybridization with perfectly matched C. reinhardtii DNA, there would
be many features that would consistently produce a much weaker or
absent signal with S1D2 because of polymorphisms. Once such
differences are identified, they become mapping markers. Given the
known frequency of polymorphisms between S1D2 and reinhardtii, there
are probably hundreds of thousands of such markers that could be
scored by a single hybridzation on a WGA.
The WGA can be used for rapid mapping of point mutations. Mutant
progeny isolated from a cross between C. reinhardtii (the parent with
the mutation) and S1D2 are separated and pooled (the more individuals
the better). A single prep of pooled mutant DNA is hybridized to the
WGA. Any polymorphic array features that are linked to the mutant
locus will stand out as being enriched for the reinhardtii allele.
The result is a bell-shaped curve of reinhardtii-specific
hybridization whose apex represents the most closely linked
feature(s) on the chip. If you are lucky the exact location of the
mutation will be revealed as a novel polymorphism. Besides giving an
accurate map location (likely within a few cm), the closest
polymorphisms to the mutation of interest can be further developed as
conventional markers for fine mapping.
Mapping insertional mutants and deletions:
Many insertional mutations and most any deletion could be localized
rapidly with a single hybridization of mutant DNA on the WGA.
All of the above applications for WGAs have been used successfully in
other model organisms. The Arabidopsis, worm, yeast and fly
communities have all adopted WGAs.
Cost and accessibility
Cost and accessibility are potentially the biggest barriers to the
use of WGAs. The current cost of a single chip is $300-400 plus the
expense of hybridization. Some cost savings can be derived from
reusing chips. For example, with routine experiments such as
mapping, a stripped and recycled chip is fine. Moreover, as
fabrication technology improves the cost of chips will decrease.
To use the array yourself you would need access to an array reader
and the necessary software (which is true for any microarray
experiment). Part of the future plan would involve obtaining
additional funds to subsidize the cost of chips, develop protocols,
and provide a service that could be used for you to outsource
experiments.
Why am I doing this?
Besides my own interest in using WGAs, we have access to some unique
resources at Salk that put us in a good position to develop a Chlamy
array. WGAs for Arabidopsis have been developed and used in Joe
Ecker's laboratory at Salk. They have technical expertise in using
WGAs for different applications as well as custom software that could
easily be adapted for Chlamydomonas with minimal expense.
What you need to do
If you have any interest in using a Chlamy WGA then email me and let
me know, and include your laboratory affiliation. If I can
demonstrate widespread interest from our community then I can more
easily convince Affymetrix to reduce the design and fabrication fees.
That will translate into more grant money to buy chips which could be
made available as a subsidized resource.
If you have any other questions or comments, please contact me.
P.S. I intend this message to reach everyone who works on
Chlamydomonas, but there is no complete email list available. Please
forward this to any colleagues you know who do not receive bionet
mailings.
--
Jim Umen
Assistant Professor
Plant Biology Laboratory
The Salk Institute
10010 N. Torrey Pines Rd.
La Jolla, CA 92037
Phone: 858-452-7645
Fax: 858-558-6379
http://www.salk.edu/labs/pbio-u/index.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://iubio.bio.indiana.edu/bionet/mm/chlamy/attachments/20040727/3a7f2ad0/attachment.html
