Did life exist on Mars? Other planets? With AI's help, we may know soon

A straightforward and trustworthy test for indicators of extant or past life on other planets has been found by scientists—dubbed "the holy grail of astrobiology."

A seven-person team reports in the Proceedings of the National Academy of Sciences journal that its artificial intelligence-based system separated biological samples from abiotic ones with 90% accuracy.

"According to Dr. Hazen, this standard analytical technique has the potential to completely transform the hunt for extraterrestrial life and expand our knowledge of the chemistry and origin of the first life on Earth." Before the samples return to Earth, it paves the way for the use of intelligent sensors on robotic spacecraft, landers, and rovers to look for indications of life.

First and foremost, the new test may provide light on the origins of enigmatic, old rocks on Earth, as well as potentially on samples previously gathered by the Sample Analysis at Mars (SAM) instrument on the Mars Curiosity rover. An analytical tool aboard called "SAM" (for Sample Analysis at Mars) may be used for the latter experiments.

"To find out if there are molecules from an organic Martian biosphere on Mars, we may already have data in hand. We'll just need to adjust our approach to fit SAM's protocols."

Lead author Jim Cleaves of the Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC, states that "the search for extraterrestrial life remains one of the most tantalizing endeavors in modern science."

"This new discovery has a lot of ramifications, but three main ones are as follows: First, biochemistry and abiotic organic chemistry are fundamentally different; second, we can determine whether ancient Earth and Mars were once home to life by examining their samples; and third, this new technique may be able to differentiate other biospheres from Earth, which could have a big impact on astrobiology missions in the future."

The cutting-edge analytical technique depends on more than just locating a particular molecule or collection of molecules in a sample.

Instead, using pyrolysis gas chromatography analysis, which separates and identifies a sample's component parts, and mass spectrometry, which calculates the molecular weights of those components, the researchers showed that AI can distinguish biotic from abiotic samples by identifying subtle differences within a sample's molecular patterns.

AI was trained to estimate the origin of a new sample using vast multidimensional data from the molecular analyses of 134 known abiotic or biotic carbon-rich samples. AI was able to accurately identify samples with an estimated 90% origin accuracy from:

Things that have been preserved in fine-grained rock, including living things like human hair, leaves, grains, insects, teeth, bones, and contemporary shells

samples having abiotic origins, such as pure laboratory substances (such amino acids) and carbon-rich meteorites, or remnants of past life that have been modified by geological processes (such as coal, oil, amber, and carbon-rich fossils).

The authors also note that because collections of organic molecules, whether biotic or abiotic, have a tendency to deteriorate with time, it has been challenging to pinpoint the origins of many ancient carbon-bearing materials up until this point.

Remarkably, the new analytical technique found evidence of biology survived in some cases across hundreds of millions of years, despite extensive degradation and modification.

According to Dr. Hazen, "We started with the premise that there are 'chemical laws of life' that impact the variety and distribution of biomolecules; that is, that the chemistry of life is fundamentally different from that of the inanimate universe. If we could figure out such guidelines, we could use them to direct our efforts in simulating the beginnings of life or looking for hints of life on other planets."

These findings suggest that, despite the fact that extraterrestrial life may differ greatly from that of Earth, we might be able to locate life on other planets or in other biospheres. Furthermore, we can determine if life on Earth and other planets originated from a shared or distinct source if we do discover evidence of life elsewhere."

Stated differently, both Earthly life and extraterrestrial biochemistries should be detectable by the technique. That is significant because, although it is not unreasonable to anticipate that extraterrestrial life will employ DNA, amino acids, and other biomarkers, it is not possible to identify the molecular biomarkers of Earth life. With our approach, we search for patterns in molecular distributions that result from the need for "functional" molecules in life."

We taught our machine-learning model to predict only two sample types: biotic or abiotic, thus it truly surprised us when the system found three different populations: abiotic, live biotic, and fossil biotic.

Put otherwise, it has the ability to distinguish between more recent biological samples derived from fossil samples—for example, a freshly picked vegetable or leaf against a long-dead object. We are encouraged by this unexpected discovery that further characteristics, such photosynthetic life or eukaryotes (cells with a nucleus), may potentially be identified."

Co-author Anirudh Prabhu of the Carnegie Institution for Science uses the example of sorting coins based on various characteristics—monetary value, metal, year, weight, or radius, for example—and then combining those attributes to find more complex groupings and separations to illustrate the role of AI. "And AI algorithms are invaluable to collate the data and create highly nuanced insights when hundreds of such attributes are involved."

"The distinctions between biotic and abiotic samples pertain to chemical aspects such as solubility in water, molecular weights, volatility, and other related factors," Dr. Cleaves continued.

"A cell contains a membrane and an interior known as the cytosol; the content of the cell is mostly water soluble, while the membrane is largely water-insoluble. That's how I would conceptualize things simply. This configuration both prevents the "inner components" from leaking through the membrane and maintains the membrane together in an effort to reduce the number of times its components come into contact with water."

"Inside components, such as proteins and chromosomes, can also remain dissolved in water despite their enormous size," he explains.

Thus, dissecting a live cell or tissue into its constituent parts yields a mixture of extremely water-soluble and extremely water-insoluble molecules dispersed along a spectrum. Over their lengthy histories, materials like coal and petroleum have lost the majority of their water-soluble content."

"Biological distributions are not the same as those of abiological samples, which can have distinct distributions across this spectrum with respect to one another."

The method might soon provide answers to many of the planet's scientific puzzles, such as where the black sediments from Western Australia, which date back 3.5 billion years, came from. These are controversial rocks that some scientists believe contain the oldest fossil bacteria on Earth, while others maintain they contain no indications of life.

Similar arguments are raised by other samples from ancient rocks found in China, South Africa, and Northern Canada.

"We are currently utilizing our techniques to tackle these enduring inquiries regarding the biogenicity of the organic matter within these rocks," states Hazen.

Furthermore, fresh concepts on the possible benefits of this novel strategy in other disciplines including biology, paleontology, and archaeology have proliferated.

What more insights would we be able to obtain if AI is able to discern between biotic and abiotic, as well as between current and ancient life with ease? Could we, for instance, determine if a prehistoric fossil cell was photosynthetic or had a nucleus?" asks Dr. Hazen.

Could it distinguish between several types of wood from an archaeological site based on its analysis of charred remains? It seems as though we are only beginning to scratch the surface of an enormous ocean of opportunities."

"Students of Earth's early history may find Cleaves and colleagues' inventive technique for separating abiotic organic matter from living matter to be a gift for astrobiologists. Although much remains to be discovered, one day a next-generation version of their system may very well travel to Mars to assess the likelihood of life there while its Earthbound sisters demonstrate the age of life on Earth," stated Andrew H. Knoll, Fisher Research Professor of Natural History and Research Professor of Earth and Planetary Sciences Emeritus, Harvard University Department of Organismic and Evolutionary Biology

"This new study intrigues me greatly! As it seems to separate abiotic from biotic organic matter based on its molecular complexity, it is a novel line of inquiry to pursue and might prove to be an invaluable instrument for astrobiology missions, stated Emmanuelle J. Javaux, Head of the Early Life Traces and Evolution-Astrobiology Lab and Director of the Astrobiology Research Unit at the University of Liège in Belgium.

Testing this novel technique on ancient and extant creatures from the three domains of life as well as on some of the earliest disputed and hypothesized signs of Earth life would be fascinating! This might settle some heated arguments in our neighborhood."

According to Karen Lloyd, a professor in the University of Tennessee, Knoxville's Department of Microbiology, "we are in great need of biosignatures for life that don't depend on looking for a specific type of biomolecule that may be universal to all life on Earth, but not universal to all life outside of Earth."

"This work outlines a future direction for assessing the likelihood of a chemical signature being suggestive of life or not, without assuming that extraterrestrial life will utilize the same biomolecules as terrestrial life. The range of measures that may be utilized to discover agnostic biosignatures of life may be increased by using the same statistical method to other kinds of measurements as well."

This offers a valuable prospective tool for identifying life on other planets and in the far past of Earth. Importantly, the method may already be applied on spacecraft that can explore various solar system regions in our quest for extraterrestrial life," said Daniel Gregory, assistant professor at the University of Toronto's Department of Earth Sciences.