MODELS FOR COLLABORATIONS AND COMPUTATIONAL BIOLOGY - Collaborative Computational Technologies for Biomedical Research - page 48

with private funding from several pharmaceutical company partners
(GlaxoSmithKline, Merck & Co., Novartis, and Pfi zer) and is managed by
the FNIH.
The overall goal of the OAI is to generate an unparalleled, state-of-the-
art database that is capable of explaining the natural progression of osteoar-
thritis and providing relevant information on imaging and biochemical
biomarkers as well as outcome measures for osteoarthritis. Originally designed
as a seven-year project, the success of this consortium led to a second genera-
tion of the OAI, which is presently under development. For more detail
see http://oai.epi - ucsf.org/datarelease/ and http://www.niams.nih.gov/Funding/
Funded_Research/Osteoarthritis_Initiative/default.asp .
3.5 OPPORTUNITIES FOR COMPUTATIONAL BIOLOGY
RESEARCH PARTNERSHIPS
While some areas of research have been quicker to embrace the use of large-
scale partnerships to advance individual research interests, the fi eld of com-
putational biology seems to be slow in warming up to this new model. As a
result, to date the most successful large collaborations in computational
biology have been run as “ divide - and - conquer ” strategies, as opposed to a
“ too - many - cooks - in - the - kitchen ” scenario of shared analyses. As such, existing
collaborative tools in our experience do not go much beyond teleconferencing,
screen sharing/multicasting (e.g., via Virtual Network Computing [VNC]), and
document editing (e.g., wikis). Over the last two decades, the accelerating pace
of technological progress has generated a massive volume of biomedical data,
opening a new era of life science investigation. For example, recent technical
advances in the fi eld of microarrays have produced considerable quantities of
gene expression and metadata associated with various human diseases and
conditions. The ongoing evolution of new technologies required computa-
tional methodologies to evolve in parallel.
As a result of this evolution, collaborative initiatives in computational
biology have started to emerge to address the unique challenges arising in the
fi eld. In particular, an area that received special attention was the establish-
ment of standards for data storage and data management. An illustrious
example is given by the Minimum Information About a Microarray Experiment
(MIAME) consortium, which outlines the content and structure of the neces-
sary information required for recording and reporting microarray-based gene
expression data [8]. The guidelines provided by MIAME include the standards
for the raw data (e.g., Affymetrix CEL or GenePix GPR fi les), the processed
(normalized) data, the sample annotations, the experimental design, the anno-
tation of the array, and the laboratory and data processing protocols [8].
Similar examples of collaborative efforts to develop and agree on common
vocabularies and standards include the International Health Terminology
Standards Development Organization, which developed, maintains, and pro-
motes suitable standardized clinical terminologies, notably SNOMED [9].