Biomedical Engineering Reference
In-Depth Information
14.8 ADDITIONAL BENEFITS OF CLOUD COMPUTING
FOR SYSTEMS BIOLOGY
Cost and convenience are the other drivers in the move toward cloud comput-
ing in the sciences. Specifying, purchasing, and confi guring a traditional
compute cluster is an enormous undertaking of time and money. Operating
and maintaining a cluster is no easier: Once installed, nodes begin to fail
almost immediately, obsolescence begins its steady creep, and even under ideal
operating conditions the cluster is almost never optimally used—it is either
under capacity or maxed out. On the other hand, a cloud-based cluster requires
no maintenance, requires no up-front money or planning, can be made just as
large or small as is needed in the moment, and costs nothing at all when not
actually in use. Compute power becomes more like a utility or reagent and
frees up chunks of capital for more interesting lab equipment.
14.9 SOME EXAMPLES OF CLOUD - BASED
SYSTEMS BIOLOGY TOOLS
Even when collaboration is not a goal, access to a system already confi gured
with useful tools can save hours or days of setup efforts. Amazon EC2 provides
online access to familiar Linux and Microsoft Windows setups (Amazon
Machine Images, or AMIs) that can be readily customized, then saved for
use by others, and has become a hotbed of helpful preconfi gured images. Here
is a small sample of useful Amazon EC2 machine images, all of which are
publically available and cost about 10 cents (U.S.) per hour (paid to Amazon)
to run:
• ViPDAC ( http://proteomics.mcw.edu/vipdac ) provides a ready - to - run
EC2 image of proteomics tools, including BLAST, OMSSA, and
X!Tandem.
• CycleCloud for Life Sciences ( http://my.cyclecloud.com/info/lifesciences/ )
provides EC2 implementations of many popular tools, including BLAST,
Bowtie, X!Tandem, OMSSA, R, and many others. There is a paid option
that includes support and extra features, but the standard offering costs
you only what EC2 charges for the use of its compute nodes.
• BioConductor ( http://www.bioconductor.org/ ) EC2 images maintained
by Martin Aryee at Johns Hopkins University: http://www.biostat.jhsph.
edu/
maryee/index.php/Cloud/BioconductorAMI .
• J. Craig Venter Institute ' s BioLinux ( http://www.jcvi.org/cms/research/
projects/jcvi - cloud - biolinux/overview ) provides EC2 implementations of
BLAST, glimmer, hmmer, phylip, rasmol, genespring, clustalw, the Celera
Assembler, and the EMBOSS collection of utilities.
Others are readily found by fi ltering on “ bio ” in Amazon ' s list of public images.
Search WWH ::




Custom Search