Welcome (original text of announcement)
Dent
this email from David, but for posterity and for any members who heard
about the project from other source I'm posting it here.
d
On Dec 20, 2011, at 9:33 AM, David Haussler wrote:
Dear colleagues,
I'm writing to you to announce a collaborative project in whole-genome
alignment (wga) assessment organized by my student Dent Earl, and to
ask you to participate.
As you're probably aware our lab is deeply involved in the Genome 10K
project (G10K), an international project to sequence the genomes of
10,000 vertebrate species. This project has already spawned a
successful spin-off project, the Assemblathon, a collaborative
competition to assess the state of de novo assembly[1]. As G10K ramps
up and more and more whole-genome assemblies are produced, one of the
first questions that will be asked is, "How are these species, these
genomes, related?" We think this glut of data will lead to a
resurgence of interest in alignment in general and wga in particular.
With that near future in mind, our group has decided the time has come
for an Assemblathon of alignment, an "Alignathon," if you will. We
have commitments from several established wga pipelines, but we'd like
to include the broader community. The goal is find the best aligners
for use on the G10K data set and for the large scale projects of the
near future.
Which is why I'm writing; I would love for you to participate.
The test suite my team has designed is broken into three parts: two
simulated sets (a four- species primate-like phylogeny and a five-
species mammal-like phylogeny) created using the Evolver simulation
suite[2] and one set comprising the 12 fly genomes. All the genomes in
the sets are approximately 120-200 Mb with between four and seven
chromosomes. We hope that using genomes 1/10th the size of a typical
mammal genome will, like in the Assemblathon, allow groups that are
developing new methods (which may not yet be able to scale to full
genome sizes) to still participate. To unite these sets, my team has
created a testing framework that is both simple and extensible so that
each participating group may run analyses independently and repeatedly
and can submit their own evaluations to the project. Just like in the
Assemblathon, participating groups will be included in the authorship
of the paper that comes out of the project.
We're shooting for a final alignment submission deadline of Friday,
February 24. If you are interested, you can read more about the
project on the website Dent has set up:
http://compbio.soe.ucsc.edu/alignathon/
If you would like to join the project please join the google group:
http://groups.google.com/group/alignathon
General questions can go to Dent Earl <de...@soe.ucsc.edu>.
Best regards,
David
[1. Earl et al. 2011 Assemblathon 1: A competitive assessment of de
novo short read assembly methods. Genome Research 21(12): 2224-2241
http://genome.cshlp.org/content/21/12/2224 ]
[2. Evolver Edgar RC, Asimenos G, Batzoglou S, Sidow A. http://drive5.com/evolver/
]