Personalized health care solutions demand an integrative, systems-level approach to the understanding of human biocomplexity. Research on the microbiome is a core component of that approach (Nicholson, 2006) (Box 2-1). Genes are just one component of the gene-diet-microbial interactions that make humans the “super system” they are. So while genome-wide association studies (GWAS), for example, are very popular, they are not always, in Jeremy Nicholson’s words, “tremendously revealing,” because statistical significance often has very little to do with biological significance. Speliotes et al. (2010) reported having found 32 statistically significant body mass index (BMI)-associated genes; in fact, the BMI-linked genes accounted for only 1.47 percent of the variance in BMI. Nicholson said, “It is all the other stuff in the world that is really important with the BMI.” He mentioned work by Jeffrey Gordon and colleagues showing a connection between obesity and the microbiome (Ley et al., 2006; Turnbaugh et al., 2006). While the obesity-microbiome connection is controversial, Nicholson pointed out that the mixed results driving the controversy are due to differences in levels of phylogeny and in the way the experiments have been conducted.16
This is not to say that systems-level studies of human genome complexity are not generating interesting information. They are. For example, Loscalzo et al. (2007) used a systems-level approach to show that almost all human diseases are genetically connected, with the same gene(s) being implicated in different disorders. However, understanding the human genome itself is not enough if the vision of personalized, or “precision,” health care is to be realized (Mirnezami et al., 2012). The microbiome represents yet another entire level of genetic connectivity. The challenge for the future, Nicholson said, is “to think about layers of networks on top of networks.” That is, how does the human genome network(s) interact with the microbiome genome network(s), across both time and space? “This is really quite a tough problem,” Nicholson said, “probably the toughest problem in 21st-century biology.”
The Metabolic Window on Complex System Activity
Complex interactions between the microbiome and its host generate more than differential disease risks. They also generate differential metabolic phenotypes (Holmes et al., 2008a). In fact, disease risks and metabolic phenotypes are both biologically and statistically linked such that one can
15 This section summarizes Jeremy Nicholson’s presentation.
16 See the summary of Peter Turnbaugh’s presentation in Chapter 3 for a more detailed discussion of evidence suggesting an obesity-microbiome connection.