University of Colorado Boulder Ph.D. student Corinne Walsh works with soil samples containing microbes associated with wheat plants. Credit: Cooperative Institute for Research in Environmental Sciences (2020)
Key points:
- Using machine learning, researchers developed a system to quickly predict bacteria’s environmental pH preferences.
- The work could guide ecological restoration efforts, agriculture, and the development of probiotics by giving researchers a solid starting guess for which pH to use.
- The team plans to expand their research by exploring temperature preferences of bacteria.
Machine learning is helping researchers figure out a way to predict bacteria’s environmental pH preferences by studying their genomes. Experts at the University of Colorado Boulder developed a new approach that promises to help guide ecological restoration efforts, agriculture, and even the development of health-related probiotics.
Understanding if certain bacteria is likely to thrive in acidic, neutral, or basic environments can help researchers anticipate how microbes will adapt to almost any environmental change.
For the work, published in Science Advances, the research team used what scientists already know about selected bacterial groups that prefer one particular pH over another, and then used machine learning to link those groups’ environmental pH preferences with their genetic makeup. The work involved sorting through the genomes of more than 250,000 types of bacteria derived from close to 1,500 lake, soil, and stream samples.
“What we found is we can make inferences about their pH preferences based on genomic data alone,” said lead author Josep Ramoneda, a CIRES visiting scholar at the University of Colorado at Boulder.
Normally, it’s next to impossible for scientists to isolate and grow microbes in the lab. But these findings can help researchers grow colonies of finicky bacteria by giving them a first guess at what pH to use. Rather than it taking possible years to figure out how to “culture” bacteria, the machine learning-based method can simplify and speed the process.
The new data can also give agricultural and forestry experts a leg up. Typically, they add live bacteria to “inoculate” growing plants with helpful communities of bacteria. Now, they may get quicker, better insight into the types of bacteria that might help restore a native prairie vs. pine forests or better grow corn or soybeans by ensuring that inoculants will be adapted to the local pH
Next, the team says they will explore the temperature preferences of bacteria, another complex system likely involving many genes. This could help them better understand how an increasingly warm world will influence soil bacterial communities.