Cancer researchers are embarking on a huge experiment that is attempting to deliver on science’s promise to usher in the era of personalized medicine. Led by the National Cancer Institute, the project is called the Cancer Genome Atlas. The idea is to catalog all the genetic mutations associated with cancer. The positive spin on this project is that it’s highly ambitious, but some have called it foolhardy. Tonight, a room of brain cancer researchers hashed over their portion of the Cancer Genome Atlas – a pilot project to catalog genetic mutations in one form of brain cancer called glioblastoma multiforme. And, as scientists are wont to do, they spent more than an hour pointing out all the flaws with the design of the brain cancer part of the atlas.

They weren’t just trying to be difficult; this is part of what scientists are trained to do. Their job is to study things that, by nature, are often very poorly understood. So they have to eliminate as much uncertainty as possible from the outset, because the chances of getting a useful answer back to any of their questions is often very small. So, tonight, when asked to brainstorm about the Cancer Genome Atlas project, the room full of somewhere around one hundred scientists put their collective minds to work pointing out all its unanticipated shortcomings. Had its leaders chosen the right brain cancer cells to study? Had they made a plan for analyzing the data? And how, exactly, were they going to convince all the brain cancer researchers to work together?

This last question may seem odd, because to outsiders, brain cancer researchers probably just seem like a group of individuals dedicated to one common goal: curing brain cancer. But it’s not that simple. Brain cancer researchers are actually a group of diverse little cliques, each of which works on one tiny part of the total problem. So there are basic lab scientists who study cancer cells in Petri dishes, and never see a live, human patient; and there are neurosurgeons, who cut brain cancers out of sick people and send them off to the lab for analysis; and lots of other types in between. All too often, these groups of scientists will never speak to each other. So a lot of what the scientists were really debating tonight was: how do you propose to get those guys, or gals, to work with us?

Finally, Paul Mischel of the University of California at Los Angeles injected a much-needed reality check into the discussion. It shouldn’t be that hard to get people to work together, he said: “Let’s focus on thoroughness – on taking two or three drug targets all the way to the clinic to treat patients.” If every scientist in the room brought his or her expertise to bear on a project involving one of these common drug targets, he was saying, then tight-knit, focused research groups would form naturally. After all, he said, the whole point of the Cancer Genome Atlas is to get new treatments out to patients – and if everyone unifies around that goal, certainly, it shouldn’t be so hard to work together.

It was refreshing to hear a scientist bring the discussion back to the real world – I only hope some of the people in the room were listening.