Researchers find via machine learning that differentiating decision-making components for retardation, growth, and saturation phases of bacterial population growth protects the population from extinction

Microbial populations may be small but they are surprisingly complex, making interactions with their surrounding environment difficult to study. But now Japanese researchers have discovered that machine learning can provide the tools to do just that. In a study published this month in eLiferesearchers from the University of Tsukuba have revealed that machine learning can be applied to the growth of bacterial populations to discover its link to variations in their environment.

The dynamics of microbial populations are generally represented by growth curves. Typically, three parameters taken from these curves are used to assess how well microbial populations adapt to their environment: lag time, growth rate, and saturated population size (or carrying capacity). These three parameters are probably related; trade-offs have been observed between growth rate and lag time or population size within species, and with related changes in saturated population size and growth rate among genetically diverse strains.

“Two questions remain: are these three parameters affected by environmental diversity, and if so, how?” says the study’s lead author, Professor Bei-Wen Ying. “To answer this, we used data-driven approaches to study the growth strategy of bacteria.”

The researchers constructed a large dataset reflecting the dynamics of Escherichia coli populations under a wide variety of environmental conditions, using nearly a thousand combinations of growth media composed of 44 chemical compounds under controlled laboratory conditions. They then analyzed big data for relationships between growth parameters and media combinations using machine learning (ML). ML algorithms built a model based on sample data to make predictions or decisions without being specifically programmed to do so.

The analysis revealed that for bacterial growth, the decisional components were distinct between the different growth phases, for example, serine, sulfate and glucose for growth retardation (lag), growth rate and growth maximum (saturation), respectively. Results from further simulations and analyzes showed that branched-chain amino acids likely act as ubiquitous coordinators for bacterial population growth conditions.

“Our results also revealed a common and simple strategy for risk diversification under conditions where bacteria experienced excess resources or starvation, which makes sense in both an evolutionary and ecological context,” says Professor Ying.

The results of this study revealed that exploring the world of microorganisms with data-driven approaches can provide new insights that were previously inaccessible via traditional biological experiments. This research shows that the ML-assisted approach, although still an emerging technology that will need to be developed in terms of biological reliability and accessibility, could open new avenues of applications in the life sciences, in particular microbiology. and ecology.

The study was funded by the Japanese Society for the Promotion of Science 21K19815 and 19H03215.

Source of the story:

Material provided by University of Tsukuba. Note: Content may be edited for style and length.