Oregon State University scientists have postulated a microbiological version of the "Anna Karenina principle” in perturbed microbiotas, suggesting, as Tolstoy did, that "All happy families are alike; each unhappy family is unhappy in its own way."
"When microbiologists have looked at how microbiomes change when their hosts are stressed from any number of factors, temperature, smoking, diabetes, for example, they've tended to assume directional and predictive changes in the community," said Rebecca Vega Thurber, corresponding author on the perspective study funded by the NSF and published yesterday in Nature Microbiology. "After tracking many datasets of our own we never seemed to find this pattern but rather a distinct one where microbiomes actually change in a stochastic, or random, way."
Lead author Jesse Zaneveld, of the University of Washington-Bothell, collaborated with Vega Thurber and her student Ryan McMinds, to survey the literature on microbial changes caused by perturbation. Together they found those stochastic changes to be a common occurrence, but one that researchers have tended to discard as "noise" rather than report
"Thus we present the Anna Karenina principle for microbiomes," Vega Thurber said. "When microbiomes are happy they are all similar in their composition but during stress or unhappiness they change in a multitude of distinct ways. This piece draws together diverse microbiome research. We think this is an important emerging paradigm for thinking about microbiome data. We present ways of identifying it and distinguishing it from other patterns."
Studies of microbiome dynamics have typically looked for patterns that shift microbiomes from a healthy stable state to a dysbiotic stable state.
"The Anna Karenina principle is a complementary alternative," Vega Thurber said. "The changes induced by many perturbations lead to transitions from stable to unstable community states, dysbiotic individuals vary more in microbial community composition than healthy individuals."
"Our message to researchers is, don't throw out these observations as noise, but include this principle in the microbiome pipelines and software so that scientists can press a button that gives you the answer to, 'Do I see the Anna Karenina principle in the dataset,'" Vega Thurber said.