(Author's note: It was recommended by a loyal reader that I bring more sports into my blog. Since the name of the blog is PolITiGenomics and not AthlITigenomics, this is as close as you are going to get.)
This week the NBA finals tip off. The match up is a throwback: Celtics and Lakers. The last time these two teams played each other in the finals was in the 1980's. Then all the talk was about Magic Johnson and Larry Bird. But there was another interesting character on the court at that time, Kurt Rambis. Mr. Rambis was basketball's quintessential 6th man (although at points, he was a starter for the Lakers). The guy who comes off the bench when the starter needs a rest or gets in foul trouble, when you need a little extra defense, or when you just need to change the momentum. He wasn't a highly skilled player, but what he lacked in skill he made up for in hustle.
Biomedical research's 6th man is the informatician. Typically she is not called in until the 3rd quarter when things are headed in the wrong direction and there is not much time to turn things around. The skills of the informatician are often not fully appreciated by the rest of the research team or the fans (funding agencies and manuscript reviewers), but her skills are useful and sometimes even necessary. With the advent of genome-wide association studies and massively parallel sequencing, that "sometimes" in the previous sentence is becoming an "always".
A couple weeks ago I attended a meeting at The Ohio State University about "The Role of Biomedical Informatics in Overcoming Barriers in Cancer Research" that was hosted by the NCI and OSU. In the research community, there is an increasing awareness that informatics needs to play a larger role in experimental design, not just post-experiment analysis. The clinical community is not so sure. Indeed, even the research community seems a bit perplexed about how to do this. Most discussions of informatics or bioinformatics these days deals with managing and categorizing the massive amounts of data being generated. There is not much discussion about how informatics can help to improve experimental design, from ensuring the clinical samples have the proper descriptive data associated with them (annotation) to specifying how many samples you need. Not to mention it is a bit self serving for informaticians to suggest to a group of researchers that their experiments would be a lot better if they worked more closely with the informaticians.
Another barrier is the culture. Informaticians think in terms of cohorts, samples, databases, and web portals. Clinicians think in terms of patient outcomes. Clearly both approaches can be focused on the same goal, but the starting point of each discipline is worlds apart. A strategy for bringing these two worlds together was hinted at by a question asked by Dr. Larry Norton from Memorial Sloan-Kettering Cancer Center, "Can biomedical informatics provide information that helps doctors treat patients?" Ken Buetow of NCICB put it another way: in the same way that people who eat sausage do not know (and probably don't want to know) how sausage is made, doctors and patients do not care how the data was analyzed to allow for better treatment, they just want better treatment. Anna Barker, Deputy Director of NCI, brought these comments (and much of the meeting) together with her charge at the opening, "Biology is a circle. Medical bioinformaics need to become a part of that circle, an organic part." Over the coming days and weeks I will present some of the ideas from the meeting for making medical bioinfomatics an organic part of biology and clinical research.