Using Machine Learning to Create a Bacterial Evolutionary Timeline

Using Machine Learning to Create a Bacterial Evolutionary Timeline

a comprehensive method to reconstruct the history of oxygen adaption and date bacterial evolution. By combining genomic, fossil, and geochemical data and connecting aerobic metabolism and oxygen tolerance to the GOE, we were able to construct a bacterial timetree. Anaerobic (blue) and aerobic (red) states are indicated by colors, while the percentage of aerobic lineages in the current bacterial phyla is shown by shades of purple. Plasmids and mitochondria were added in order to take advantage of the larger eukaryotic fossils. For the purpose of time comparison, land plants and animals are indicated. — Science

Scientists from the University of Queensland have contributed to the creation of a thorough timeline for the evolution of bacteria, indicating that some bacteria consumed oxygen long before they developed the capacity to manufacture it through photosynthesis.

Researchers from the Okinawa Institute of Science and Technology, the University of Bristol, Queensland University of Technology, and UQ led the multinational collaboration, which examined how microorganisms reacted to the Great Oxygenation Event (GOE), which occurred approximately 2.33 billion years ago and transformed Earth's atmosphere from one that was primarily oxygen-free to one that is suitable for human breathing.

The lack of complete fossil evidence has made it challenging to date the evolution of bacteria before, during, and after the GOE, according to Professor Phil Hugenholtz of UQ's School of Chemistry and Molecular Biosciences.

According to Professor Hugenholtz, "the bulk of microbial life leaves no direct fossil record, which means that fossils are missing from the majority of life's history on Earth."

However, by simultaneously examining geological and genomic data, we were able to fill in the gaps since we know that ancient rocks include chemical clues about how bacteria lived and nourished.

Assuming that most aerobic bacterial branches are unlikely to be older than this event unless genetic or fossil indications indicate otherwise, the GOE was used as a time boundary. This was the main innovation.

Initially, the group calculated which genes were found in ancestral genomes. Then, they employed machine learning to forecast if each ancestor relied on oxygen for survival.

The researchers incorporated genes from mitochondria (related to alphaproteobacterial) and chloroplasts (related to cyanobacteria) into fossil records in order to make the greatest use of them. This allowed them to use information from early complex organisms to more accurately determine the time of events.

"The findings indicate that at least three aerobic lineages existed around 900 million years prior to the GOE, indicating that the ability to use oxygen arose long before it was widely accumulated in the atmosphere," Professor Hugenholtz stated.

"There is evidence that the cyanobacterial progenitor underwent the earliest aerobic transition approximately 3.2 billion years ago, suggesting that aerobic metabolism took place prior to the evolution of oxygenic photosynthesis."