CANCER COMMONS: BIOMEDICINE IN THE INTERNET AGE - Collaborative Computational Technologies for Biomedical Research - page 166

outsourced, but these usually require extensive (and expensive) management
and information infrastructures. The Internet has upended the economics
of this sort of virtualization, bringing enterprise management tools such
as SalesForce.com (http://www.salesforce.com/), computational tools such as
BioBike (http://www.biobike.org/ and [3]), databases such as those found at
the National Center for Biotechnology Information (NCBI, http://www.
nbci.nlm.nih.gov/), and enterprise information infrastructure such as Google
Apps (http://apps.google.com/), well within the range of price and usability of
small distributed teams. Moreover, the technologies required for drug devel-
opment, such as high-throughput sequencing, drug screening, computational
modeling, and so forth, have traditionally been available only inside large
pharmaceutical and biotech companies. However, rapid technological advances
have made such services readily available at reasonable cost. Indeed, noncor-
porate virtual treatment discovery projects are already emerging, funded by
nonprofi t disease foundations like the Cure Huntington's Disease Initiative
(http://www.highqfoundation.org/), the Myelin Repair Foundation (http://
www.myelinrepair.org/), the Michael J. Fox Foundation for Parkinson's
Research (http://www.michaeljfox.org/), and others.
There are several unique aspects of the way that virtual biotech is deployed
in the setting of Cancer Commons. First, personalized virtual biotechs can be
created out of Cancer Commons when no specifi c treatment hypotheses can
be found in the molecular disease model for a particular patient's disease. The
molecular disease model is in this sense a cache of previously solved targeted
drug discovery activities: A particular patient seeking information regarding
potential treatments queries the model using his or her clinical and genomic
profi le. If there is information regarding one or more potential treatments
already cached in the model, they are returned as treatment hypotheses , along
with an explanation of why they were chosen. Signifi cantly, these treatment
hypotheses can include available therapies that proved effective on patients
with other types of cancer whose tumors exhibited similar genomic or path-
way aberrations. If, on the other hand, no such hypotheses are forthcoming,
then a personalized virtual biotech project can be created to try to fi nd a
treatment.
The second unique feature of personalized virtual biotech, as it operates in
the Cancer Commons setting, is that the Cancer Commons platform supports
the process via an integrated services and workfl ow architecture, integrated
state-of-the-art bioinformatics, and, most importantly, the ability to run real-
time live in-patient experiments where real patients are treated by real physi-
cians and their genomic and clinical profi les and their response to treatment
are tracked. These factors enable researchers involved in a personalized virtual
biotech to hypothesize and test novel treatments in settings ranging from in
silico to in patient. This will be explained in more detail in the next section.
Historically, good leads are often found by using existing drugs off-label or
combining them in cocktails. This strategy has the advantage that these leads
can be used immediately in patients. The Cancer Commons platform can