How does a cell
transform itself into a community of communicating cells with a variety
of specialized functions? That central question of development has
acquired renewed impetus with the availability of comprehensive genomic
measurement technologies. One of the hypotheses I have been pursuing
is that many diseases, not only of childhood, have their roots in variations
in the developmental process and that understanding these variations
will allow us to effect earlier diagnoses and ultimately targeted therapies.
The range of disease is from cognitive disorders such as autism to
several forms of malignancy. The principal approach I have taken is
the study of the dynamics of genomic physiology at developmental time
scales to obtain more mechanistically-based and informative models.
This interest in explanatory dynamic (often Markovian) models dates
from my early interest in population
genetics and continues to my current focus on neurodevelopment.
Given the large numbers of well phenotyped samples that are required for most of these studies, I help found the i2b2 National Center for Biomedical Computing, funded by the NIH to see how we could use the informational byproducts of healthcare delivery to drive discovery driven research. The free and open source software we developed is now being deployed at over a dozen large academic healthcenters
Personal Perspectives
An early (2002) view of the excitement and challenge and pragmatics of microarray
applications to medicine: Microarrays for an Integrative
Genomics
|
Why is it easier to find out what experience shoppers worldwide have
had with the latest digital camera than it is to determine what adverse
events patients have had with a particular drug? Why are blood tests
and X-rays repeated needlessly? If so many companies maintain our credit
history, why does the healthcare system have so much trouble maintaining
records of our allergies? These questions have generated a whole array
of responses ranging from the sociological to the economic and the
technical. However, many large engineering efforts have a similar set
of challenges. And although this kind of infrastructural engineering
work has less glamour than the scientific agenda, it is a necessity
to effect a better healthcare system. For this reason, in parallel
with my research efforts, I have worked on designing and engineering
novel health information systems over the last 15 years.
The essence of the challenge continues to be that (to my initial
chagrin) success in clinical informatics depends on mastery of the
sociological surround of the healthcare system. This has informed me
when I led the technical architecture of distributed
regional multi-hospital information systems starting in 1994 where
the design ensured that there would be no central database, only pointers
used to dynamically generate views of the longitudinal medical record.
It also informed the architecture of the PING
personally controlled medical record (rebranded as indivo) which leverages the patients’ right
to a copy of their own record and their (our) self-interpowerest in
maintaining transparency in the disposition and flow of their personal
health information, particularly in the genomic era. It has
also informed the design of the peer-to-peer
tissue access repository for use in genomic and other studies which
is currently in pilot phase in centers across the US where each tissue
bank or registry or pathology system maintains full and autonomous
control of its own data while exposing these nationwide. Finally,my concerns in the effects
of an understandable but overreaching paternalism has led to the following sytem: the "Informed Cohort"
|
