Flickr/Max Braun

Consider Manhattan. The borough is divided into neighborhoods known for certain attractions -- many of the city's best art museums are clustered on the Upper East Side. For great theater? Go to Broadway (or off-Broadway or off-off Broadway).Visit Central Park for some time outdoors. To know the city, you can't just know the individual neighborhoods, you have to know how people and things flow through the constellation of more-or-less self-regulating systems.

Understanding disease is a bit like getting to know New York, Albert-Lázlo Barabási argued in a talk today at the TEDMED conference in Washington, DC. Barabási is a physicist by training who got into studying disease by first examining networks. He believes that the way we currently practice medicine -- identifying a diseased part of the body, then working with a specialist to treat the illness that ails the organ in question -- is too specific. It doesn't account for the complex relationships between parts of the body that make up a larger system. Right now, we're focusing too much on individual buildings and neighborhoods, rather than examining the links between things.

With recent advances in genomics, researchers have access to reams of genetic code -- a kind of map of human existence. Unfortunately, reading a map requires as much knowledge about the relationships between places as it does with knowing all the individual street names. At the moment, epidemiologists may know all the street names -- i.e. the genes -- in the human genome, but they don't yet understand how all the streets and neighborhoods fit together.

Using the idea of Manhattan as a metaphor for cellular and genomic mapping, Barabási told attendees at TEDMED that organ-centric medicine will soon give way to something called relational or "network" medicine. In this view, disease is simply the end result of genes interacting with one another in certain ways. By finding out not only which genes and diseases are linked but also how different combinations of genes work to produce related illnesses, Barabási expects medicine to evolve into a kind of systems science that it isn't today.

How will we get there? Well, if researchers can map all the genes related to asthma, Barabási said, they can find other genes that interact with asthma and extrapolate for other diseases, such as chronic obstructive pulmonary disease, a common affliction among smokers. Eventually, the jumbled collection of street names will begin to make sense as distinct neighborhoods with arteries running in between.

"Learning how diseases overlap and how they relate to each other is really crucial," Barabási said. "We should stop training cardiologists and neurologists ... Future doctors will become network-ologists."

We want to hear what you think about this article. Submit a letter to the editor or write to