New Algorithm Pushes Back Evidence of Oxygen-Producing Life by Over 1 Billion Years

Machine learning breakthrough allows scientists to detect subtle chemical signatures of ancient lifeforms in rock samples

Published on Feb. 6, 2026

Got story updates? Submit your updates here. ›

Researchers have developed a machine learning algorithm capable of identifying chemical evidence of oxygen-producing life forms in rock samples dating back over 3.3 billion years - over 1 billion years older than the previous earliest known specimens. This breakthrough could rewrite our understanding of life's origins on Earth and guide the search for extraterrestrial life.

Why it matters

Studying the origins and evolution of life on Earth is a constant challenge due to the fragmented and difficult-to-interpret nature of the ancient fossil record. This new algorithm provides a powerful tool to detect even the faintest molecular signatures of ancient life, pushing back the timeline of detectable oxygen-producing organisms by over 1 billion years.

The details

The key to the algorithm's success is its broad training on a diverse range of chemical signatures, from modern organisms to organic molecules found in meteorites. This allows it to recognize patterns indicative of life, even in highly degraded samples. The algorithm can now determine the presence of life in a sample with 90% accuracy, opening up new possibilities for understanding Earth's deep past and guiding the search for life on other planets like Mars.

  • The algorithm has identified evidence of oxygen-producing life dating back 2.5 billion years.
  • The algorithm has detected biological signatures from 3.3 billion years ago.

The players

Katie Maloney

An assistant professor at Michigan State University who explains that this new technique helps read the deep time fossil record in a new way and could guide the search for life on other planets.

Got photos? Submit your photos here. ›

What they’re saying

“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.”

— Katie Maloney, Assistant Professor, Michigan State University

What’s next

The same algorithm could be instrumental in the search for life on other planets, particularly Mars. By analyzing rock samples collected from the Martian surface, scientists could potentially identify evidence of past or present microbial life.

The takeaway

This breakthrough in leveraging machine learning to detect subtle chemical signatures of ancient life represents a significant advancement in our understanding of life's origins on Earth. The potential applications extend to the search for extraterrestrial life, as the same techniques could be applied to analyze samples from other planets like Mars.