Now researchers addressed this challenge by using a new technology, called metabolomics. They measured the levels of more than 250 biochemical compounds in over 60 metabolic pathways, including lipids, sugars, vitamins, amino acids and others in blood from over 2,800 individuals. They then combined this dataset with information on more than 600,000 genetic variants (SNPs) that were detected in the genes of each of the study participants. Most of the SNPs were located in genes known to encode proteins involved in the relevant metabolic pathways. Fifteen of the SNPs had previously been associated with metabolism-related conditions, such as cardiovascular disease, kidney disease, gout, diabetes, gastrointestinal diseases, cancer and adverse drug reactions. But the new findings also uncovered a wealth of new associations that link the genetic makeup of a person to his or her biochemical capacities. This data is publicly available in an online database, accessible at http://www.gwas.eu.
Note they used only 2,800 individuals, 600,000 SNPs in each person, and about 250 chemical measurements in each person. Imagine using hundreds of thousands of people, double the number of SNPs, some other kinds of genetic differences (e.g. deletions and copy number variations), and more measures of metabolism. Plus, throw in various measures of body shape (e.g. facial shape, height, assorted widths and lengths, condition of teeth, etc), diet, allergies, and other medical history. That scale of measurement is just a few years off at most and it will turn up many functionally significant genetic differences.
Genetic testing is getting close to telling you useful things about your metabolism.
Given the exceptional size of the dataset, the researchers prioritized the data to focus on 37 SNPs that were most strongly associated with metabolic traits, 23 of which had never been described before. The 37 SNPs had very large effects on the individuals' metabolite levels and can be considered to constitute what the authors call the "genetic basis of human metabolic individuality."
Crowd sourcing could make a big impact on this sort of research. Take the SNP testing that will be available in 2012, recruit 20,000 people to get themselves SNP tested and extensively blood tested, and get them to fill in web forms of health information. A much bigger study yielding many more useful results could be done. I'm ready to sign up...