February 6, 2026 at 3:48 pm

New Machine Learning Algorithm Helps Scientists Identify Oxygen Producing Lifeforms In Rocks Over A Billion Years Older Than Previous Earliest Known Specimens

by Michael Levanduski

Search for Life

Shutterstock

Studying any type of history is difficult because the further you go back, the harder it is to find accurate data, and that is certainly true when it comes to studying ancient life on Earth.

Scientists have been able to learn a lot about life on the planet from millions of years ago, thanks to things like fossils, specimens frozen in ice, and other techniques. When trying to learn about life on a microscopic level, however, things get even more difficult.

Researchers have found a way to look back much further than previously possible. This was done by creating a machine learning algorithm that can look at rock samples and pinpoint chemical evidence in even the smallest amounts.

According to a study published in the Proceedings of the National Academy of Sciences, this algorithm has allowed the researchers to find evidence of oxygen-producing life from samples that are 2.5 billion years old. In addition, they found biological signatures that are 3.3 billion years old.

These findings push the detection of life back by over a billion years, which is remarkable. In a statement about the study, co-author Katie Maloney, who is an assistant professor at Michigan State University, said:

“Ancient rocks are full of interesting puzzles that tell us the story of life on Earth, but a few of the pieces are always missing. Pairing chemical analysis and machine learning has revealed biological clues about ancient life that were previously invisible.”

Fossils

Shutterstock

To accomplish this, the researchers trained the machine learning algorithm on how to detect fossilized chemical signatures. This was done by feeding in modern animal and plant signatures, as well as organic molecules from meteorites.

The end result is that the algorithm is able to determine if there was life present in a sample with 90% accuracy.

This will allow researchers to look for molecular traces in rocks that are far older than was previously possible. In addition, the same algorithm could be used to look for life on Mars and other planets in the future. Maloney said:

“This innovative technique helps us to read the deep time fossil record in a new way. This could help guide the search for life on other planets.”

Having the ability to look for life in the distant past will help to unlock the many mysteries of how and when life formed on Earth.

If you enjoyed that story, check out what happened when