New Directions In Science

Draft

THREE VARIATIONS ON THE THEME

High Throughput Technologies

The effects of these is just beginning. There are 3 practical issues:

a. moving target
These technologies are evolving rapidly. The newest platform may be obsolete a year after it is introduced . It is unclear to what extent they are best served by every institution having their own vs. national service centers. In the former case, one needs a realistic budget just to keep up with what is coming and to subsidize new technologies that are always going to require support unlikely to be recovered from grants.

An endowment is needed to support such efforts and/or creation of a technology program that collabs with industry,

b, intellectual requirements for support
Many of these new technologies require cutting edge support esp from biomath types. This is esp, true for network analysis and proteomics. The elvel of wasted effort w/o such support is vast. A huge issue here is that by and large the positions in biomath have not been findable by NIH, even thought these same positions are seminal to so many folk's work.

An endowment is needed to support such efforts and/or creation of a technology program that collabs with industry,

c. even beyond instrumentation, many of the key areas require sophisticated investigators. For example, a SNP machine is useless w/o a population geneticist. Successful proteomics implies the presence o f both informatics, protein chemistry AND relevnat area sof biology.

I think this is doable by realigning young faculty searches in new directions.

Systems biology vs Systematic Biology

I make a distinction between the biology of interacting systems, the old art of "systems biology" that is dieing in many areas .. e.g try to fins a YOUNG systems physiologist, and the biolgoy growing out of the near completion of our knowledge of certain systems. I call the latter "systematic biology. "

"Systems biology" deals with the ability to use math to model complex systems ... whether that is weather or mitosis. Although some aspects of this are a new discipline, biologists have been modeling things for a long time.

I am not convinced that this as important as Lee Hood and others make it out to be because I think is generally premature. For the most part the area deals with modeling and modeling is not a dominant field in biology.

"Systematic biology" is, as I use the term, a new entity growing out of the fact that the components of several biological machines are now totally defined or soon will be. The implications for the scientific method are profound. For example one no linger asks whether we can find a protein that does something but which of the 11,000 made by one cell is the actor?

Examples of systems already at this state include:

genome sequence of man inter alia.
list of protein encoding expressed genes.
gross anatomy
taxonomy of lymphocytes and granulocytes
sub proteomes (i.e. list of proteins interacting at a reasonable affinity with one or more targets.
Genetic interactions in flies, yeast, and worms.
Many crystal structures

WHAT IS NEEDED?

All three themes share some common needs.

a. great computational support., As above this probably needs to be met from private sources for now. Uusally stats and math types end up in service like jonbs they hate. Endowing some positions do thewe folks are not servants is critical.
b. identification of biological domains ready to be jump started by the existance of systematic biology, Recruiting faculty in those domains will pay off as they will be infectious, Right now these folks are at a high premium, Harvard, UTSW, Broad, Stanford, Hopkins, UW each have a few but critical mass likely only exists at MIT.

I would argue that the recruitment is esp hard in any structure is built around traditional Depts. The key, it would seem to me is to begin with recruiting one person with the vision and then give that person some slots tied across campus to different departments. A problem here will be that by and large the able people in this field are very junior, lack the usual credentials, and are coming from model organism fields that may not be obviously relevant to biomed.

Alternatives

As ONE alternative, one can build a Biotechnology program. This is conceptually a lot easier than building a systematic biology effort, can itself solve a lot of what I said above, and often can build on existing efforts in engineering. This is exactly what Lee Hood did here. Though he and the Dean ended up fighting, it is important to note how much of that effort lives on and is functional across Seattle. This is the easiest way to go,

The SECOND , more difficult, alternative, is to build something quite ambitious that explicitly addresses the biology of one or more field that is now beginni9ng to use systematic tools. . Some of the disciplines beginning to have such people are:

immunology, nucleomics, developmental biology, protein-protein interactions, combinatorial therapeutics, pharamcogenetics

I can be more specific if you need me to and would obviously be excited to help.

Hope this helps.