Not long ago, scientists placed wagers on the number of genes in the human genome. Some bets ranged upward of 100,000 genes being present. Once the human genome sequence was completed, a project led in part by the McDonnell Genome Institute at Washington University School of Medicine in St. Louis, even the lowest guess of 25,947 proved to be above the true number.
Now, nearly 15 years later, scientists at Washington University are seeing a reminiscent trend in the newest type of big data known as metabolomics. They estimate that the number of metabolites present in a data set could be 90 percent smaller than previously estimated.
The study was published online Sept. 15 in Analytical Chemistry.
Like its genomic predecessor, metabolomics seeks to profile all of the metabolites present in a sample. Unlike genes, however, metabolites are not made from common building blocks and are much more chemically diverse. Familiar metabolites include molecules such as glucose and cholesterol, many of which are a product of diet. Thus, trying to pin down the exact number of metabolites in humans has been a tough challenge. Because of its strong nutritional dependence, some scientists have argued that it’s not even the relevant question to be asking.
There has been interest in measuring metabolites for nearly as long as there has been interest in human health. Analysis of glucose in diabetes probably dates back centuries. Handfuls of other metabolites have been used to diagnose diseases broadly referred to as “inborn errors of metabolism” since the 1960s. Metabolomics tries to measure all of these metabolites, and more. The question is: How many more are there?
The scene for metabolomics was set with the advent of sophisticated devices called mass spectrometers. These instruments are like tiny scales that can measure the weights of molecules, such as sugars. By using databases and computational algorithms, scientists can convert measured weights into compound names, like glucose.
A decade ago, when metabolomics started to become mainstream, scientists were surprised to discover that the number of signals in a typical metabolomics experiment greatly exceeds the number of known metabolites in biochemistry textbooks. Said Gary Patti, associate professor of chemistry in Arts & Sciences and senior author of the study: “Of course, the knee-jerk reaction is to assume that most of the signals that do not return matches in databases correspond to unknown metabolites that have never been reported before.”
