Researchers have announced a groundbreaking method for detecting traces of past life, which may enhance efforts to find extraterrestrial life on other planets.
Utilizing advanced chemical techniques alongside artificial intelligence, scientists have uncovering signs of ancient life in Earth’s 3.3 billion-year-old rock formations. They are optimistic that a similar methodology could be utilized on samples from icy bodies like Mars or Europa in the future.
A study published in Proceedings of the National Academy of Sciences involved analyzing over 400 samples of ancient sediments, fossils, modern flora, fauna, fungi, and meteorites to rigorously test the new detection model.
The outcome? A system capable of differentiating between remnants of life and non-living materials with more than 90% accuracy.
“This serves as a compelling example of how contemporary technology can illuminate Earth’s oldest narratives and revolutionize our exploration of ancient life on both Earth and beyond,” said Dr. Michael Wong, an astrobiologist and planetary scientist who co-authored the study. “This is a powerful new asset in the field of astrobiology.”
To extract subtle chemical signatures left by ancient organisms, the research team employed pyrolysis-gas chromatography-mass spectrometry to break down molecular structures within the samples.
Subsequently, these intricate chemical patterns were analyzed using machine learning models to identify biosignatures that were too degraded for conventional interpretation.
Co-author Dr. Robert Hazen remarked in BBC Science Focus that this technique signifies a “paradigm shift” in the field, as the algorithm does not rely on detecting specific molecules like DNA or lipids, which could indicate past life.
Instead, it focuses on the distribution of available substances and whether these patterns imply that life may have existed there.
“For the first time, we are examining distribution capabilities,” he explained. “This supports broader analyses when investigating highly degraded samples with minimal information.”
The oldest biosignature identified dates back 3.3 billion years, nearly double the previous record of around 1.7 billion years.
Additionally, researchers uncovered molecular evidence indicating that oxygen-producing photosynthesis occurred at least 2.5 billion years ago, extending the known chemical record of photosynthesis by over 800 million years.
Historically, scientists have traced life back 3.5 billion years through two main types of evidence: ancient rock formations created by microbial communities that formed sticky, layered “mats,” yielding mound-like structures called stromatolites, and observable changes in isotope ratios within the rocks.
however, suitable samples for such analyses remain rare. The new machine learning technique circumvents the requirement for intact fossils or preserved biomolecules, offering a complementary method applicable to a broader array of rocks.
The algorithm also goes beyond a basic survival or non-survival assessment. It can already differentiate between photosynthetic and non-photosynthetic organisms, as well as categorize broad cell groups known as eukaryotes and prokaryotes.
“We analyzed extensive data patterns and found clear distinctions between living and non-living entities,” Hazen noted. This capability could be vital for investigations on Mars, where scientists are uncertain about the biochemical nature of any potential life.
If retrieving samples from Mars becomes excessively costly, Hazen envisions a rover equipped with an array of devices that could apply the same machine learning technique directly on the Martian surface. His team recently secured funding from NASA to develop such an instrument package.
In the interim, the team plans to implement the technique on samples from Earth’s Mars-like deserts, aiding the groundwork for future analyses of Martian rock.
“What’s notable is that this approach does not depend on finding recognizable fossils or intact biomolecules,” emphasized co-lead author Dr. Anirudh Prabhu.
“AI has not only expedited our data analysis but also empowered us to interpret messy and degraded chemical data. AI opens new avenues for exploring ancient and extraterrestrial environments, guided by patterns we may never have considered otherwise.”
The authors cautioned that while the model is complementary to existing techniques, it should not yet be viewed as definitive proof of life. However, they believe it could become an essential analytical tool in both earth and planetary science.
“For decades, we’ve sought signs of life in ancient rocks with a limited set of tools,” remarked co-author and paleontologist Professor Andrew Knoll.
“What’s extraordinary about this work is that it enhances our toolkit and introduces entirely new, more profound questions. Machine learning can help unveil biological signals that were, until now, largely undetectable. This represents a significant leap forward in our ability to interpret Earth’s deep-time record of life.”
read more:
Source: www.sciencefocus.com
