Researchers have made significant advances in identifying some of the oldest indicators of life on our planet using a novel technique that detects the chemical signatures of ancient living organisms in rocks. This innovative approach is aimed at both unraveling Earth’s biological history and searching for life outside our planet. The research team found evidence of microbial existence in 3.3 billion-year-old rocks from South Africa, dating back to a time when Earth was approximately a quarter of its current age. Additionally, they discovered molecular signals from microbes that performed oxygen-generating photosynthesis in rocks that are about 2.5 billion years old.

Utilizing machine learning, the scientists developed a method that distinguishes between organic molecules of biological origin—derived from microbes, plants, and animals—and those of non-biological origin with over 90% accuracy. This technique is crafted to identify unique chemical patterns associated specifically with biological life.

A key contributor to this groundbreaking finding, Robert Hazen, emphasized the significance of extracting remnants of ancient life from highly degraded molecular materials. The new method involves gathering and concentrating carbon-rich molecules and analyzing them to identify thousands of small molecular fragments. Machine learning plays a crucial role in discerning patterns that differentiate ancient biological molecules from non-biological ones, a task that the human eye would struggle to perform.

Historically, scientists interested in Earth's primordial life have mostly relied on fossil evidence. Earth, estimated to be around 4.5 billion years old, likely first hosted microbial life hundreds of millions of years later in marine hydrothermal vents or terrestrial hot springs. While the oldest confirmed fossils are about 3.5 billion years old, consisting of mound-like microbial formations known as stromatolites, such remains are exceedingly rare.

The research team's new method opens up an alternative approach: examining for signs of biomolecules—substances associated with living organisms—in ancient geological layers. The method revealed molecular evidence of oxygen-producing photosynthesis being conducted by marine bacteria at least 800 million years earlier than previously recorded. This insight aligns with earlier findings that suggested Earth's atmosphere was oxygenated around 2.5 billion years ago.

Study co-lead author Anirudh Prabhu noted that this new technique effectively expands the timeline for identifying signs of life using organic molecules, nearly doubling the age capacity from 1.6 billion to 3.3 billion years. Furthermore, the method can classify various types of life forms, particularly photosynthetic organisms, and demonstrate how machine learning can identify life fingerprints in degraded ancient materials.

This cutting-edge technique has implications beyond Earth, as NASA has funded the continued development of this approach for astrobiological research. The researchers express enthusiasm about applying their method to analyze samples from Mars—either collected by current rovers or in future missions—and are also considering organic-rich plumes of Enceladus, as well as the surfaces of Saturn's moon Titan and Jupiter's moon Europa for signs of life.

In summary, this innovative research has the potential to significantly enhance our understanding of both Earth's biological past and the ongoing search for extraterrestrial life.