Carnegie Science, Washington DC
‘Holy Grail of astrobiology’: Machine learning technique reveals a sample’s biological or non-biological origin with 90% accuracy
Scientists have discovered a simple and reliable test for signs of past or present life on other planets – “the holy grail of astrobiology.”
In the journal Proceedings of the National Academy of Sciences, a seven-member team, funded by the John Templeton Foundation and led by Jim Cleaves and Robert Hazen of the Carnegie Institution for Science, reports that, with 90% accuracy, their artificial intelligence-based method distinguished modern and ancient biological samples from those of abiotic origin.
“This routine analytical method has the potential to revolutionize the search for extraterrestrial life and deepen our understanding of both the origin and chemistry of the earliest life on Earth,” says Dr. Hazen. “It opens the way to using smart sensors on robotic spacecraft, landers and rovers to search for signs of life before the samples return to Earth.”
Most immediately, the new test could reveal the history of mysterious, ancient rocks on Earth, and possibly that of samples already collected by the Mars Curiosity rover’s Sample Analysis at Mars (SAM) instrument. The latter tests could be conducted using an onboard analytical instrument nicknamed “SAM” (for Sample Analysis at Mars. (NASA photos at https://bit.ly/3P8V8II).
“We’ll need to tweak our method to match SAM’s protocols, but it’s possible that we already have data in hand to determine if there are molecules on Mars from an organic Martian biosphere.”
“The search for extraterrestrial life remains one of the most tantalizing endeavors in modern science,” says lead author Jim Cleaves of the Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC.
“The implications of this new research are many, but there are three big takeaways: First, at some deep level, biochemistry differs from abiotic organic chemistry; second, we can look at Mars and ancient Earth samples to tell if they were once alive; and third, it is likely this new method could distinguish alternative biospheres from those of Earth, with significant implications for future astrobiology missions.”
The innovative analytical method does not rely simply on identifying a specific molecule or group of compounds in a sample.
Instead, the researchers demonstrated that AI can differentiate biotic from abiotic samples by
detecting subtle differences within a sample’s molecular patterns as revealed by pyrolysis gas chromatography analysis (which separates and identifies a sample’s component parts), followed by mass spectrometry (which determines the molecular weights of those components).
Vast multidimensional data from the molecular analyses of 134 known abiotic or biotic carbon-rich samples were used to train AI to predict a new sample’s origin. With approximately 90% accuracy, AI successfully identified samples that had originated from:
- Living things, such as modern shells, teeth, bones, insects, leaves, rice, human hair, and cells preserved in fine-grained rock
- Remnants of ancient life altered by geological processing (e.g. coal, oil, amber, and carbon-rich fossils), or
- Samples with abiotic origins, such as pure laboratory chemicals (e.g., amino acids) and
carbon-rich meteorites.
The authors add that until now the origins of many ancient carbon-bearing samples have been difficult to determine because collections of organic molecules, whether biotic or abiotic, tend to degrade over time.
Surprisingly, in spite of significant decay and alteration, the new analytical method detected signs of biology preserved in some instances over hundreds of millions of years.
Says Dr. Hazen: “We began with the idea that the chemistry of life differs fundamentally from that of the inanimate world; that there are ‘chemical rules of life’ that influence the diversity and distribution of biomolecules. If we could deduce those rules, we can use them to guide our efforts to model life’s origins or to detect subtle signs of life on other worlds.”
“These results mean that we may be able to find a lifeform from another planet, another biosphere, even if it is very different from the life we know on Earth. And, if we do find signs of life elsewhere, we can tell if life on Earth and other planets derived from a common or different origin.”
“Put another way, the method should be able to detect alien biochemistries, as well as Earth life. That is a big deal because it’s relatively easy to spot the molecular biomarkers of Earth life, but we cannot assume that alien life will use DNA, amino acids, etc. Our method looks for patterns in molecular distributions that arise from life’s demand for “functional” molecules.
“What really astonished us was that we trained our machine-learning model to predict only two sample types – biotic or abiotic – but the method discovered three distinct populations: abiotic, living biotic, and fossil biotic. In other words, it could tell more recent biological samples from fossil samples – a newly plucked leaf or vegetable, say, versus something that died long ago. This surprising finding gives us optimism that other attributes such as photosynthetic life or eukaryotes (cells with a nucleus) might also be distinguished.”
To explain the role of AI, co-author Anirudh Prabhu of the Carnegie Institution for Science uses the idea of separating coins using different attributes – monetary value, metal, year, weight or radius, for example – then going further to find combinations of attributes that create more nuanced separations and groupings. “And when hundreds of such attributes are involved, AI algorithms are invaluable to collate the information and create highly nuanced insights.”
Adds Dr. Cleaves: “From a chemical standpoint, the differences between biotic and abiotic samples relate to things like water solubility, molecular weights, volatility and so on.”
“The simple way I would think about this is that a cell has a membrane and an interior, called the cytosol; the membrane is pretty water-insoluble, while the cell’s content is pretty water-soluble. That arrangement keeps the membrane assembled as it tries to minimize its components’ contacts with water and also keeps the ‘inside components’ from leaking across the membrane.”
“The inside components can also stay dissolved in water despite being extremely large molecules like chromosomes and proteins,” he says.
“So, if one breaks a living cell or tissue into its components, one gets a mix of very water-soluble molecules and very water-insoluble molecules spread across a spectrum. Things like petroleum and coal have lost most of the water-soluble material over their long histories.”
“Abiological samples can have unique distributions across this spectrum relative to each other, but they are also distinct from the biological distributions.”
The technique may soon resolve a number of scientific mysteries on Earth, including the origin of 3.5 billion-year-old black sediments from Western Australia (photo at https://bit.ly/3YWbZ4Z) — hotly debated rocks that some researchers contend hold Earth’s oldest fossil microbes, while others claim they are devoid of life signs.
Other samples from ancient rocks in Northern Canada, South Africa, and China evoke similar debates.
“We’re applying our methods right now to address these long-standing questions about the biogenicity of the organic material in these rocks,” Hazen says.
And new ideas have poured forth about the potential contributions of this new approach in other fields such as biology, paleontology and archaeology.
“If AI can easily distinguish biotic from abiotic, as well as modern from ancient life, then what other insights might we gain? For example, could we tease out whether an ancient fossil cell had a nucleus, or was photosynthetic?” says Dr. Hazen.
“Could it analyze charred remains and discriminate different kinds of wood from an archeological site? It’s as if we are just dipping our toes in the water of a vast ocean of possibilities.”
* * * * *
Comments
“Cleaves and colleagues’ innovative method of distinguishing biological from abiotic organic matter is a gift for astrobiologists and, quite possibly, for students of Earth’s early history as well. There is much still to be learned, but one day a next-generation version of their system may well fly to Mars, evaluating the possibility of life on the red planet, while its Earth-bound sisters illuminate life’s antiquity on our own planet.”
Andrew H. Knoll, Fisher Research Professor of Natural History and Research Professor of Earth and Planetary Sciences Emeritus, Department of Organismic and Evolutionary Biology, Harvard University
“I think this new study is very exciting! It is a new avenue of research to explore as it appears to discriminate abiotic from biotic organic matter based on its molecular complexity and could potentially be a fantastic tool for astrobiology missions. It would also be very interesting to test this new method on some of the oldest putative and debated traces of Earth life as well as on modern and fossil organisms from the three domains of life! This might help to solve some hot debates in our community!”
Emmanuelle J. Javaux, Head, Early Life Traces and Evolution-Astrobiology Lab, and Director, Astrobiology Research Unit, University of Liège, Belgium
“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 paper identifies a path forward for using a relatively easily measured chemical signature and determining whether it is likely to be indicative of life or not, without presuming that life outside of Earth will use the same biomolecules as life on Earth. This same statistical approach might be applicable to other types of measurements too, expanding the range of measurements that can be used to identify agnostic biosignatures of life.”
Karen Lloyd, Professor, Department of Microbiology, University of Tennessee, Knoxville
“This provides an important potential tool to identify life both on other planets and also in distant periods of Earth’s past. Importantly the technique can already be utilized on spacecraft that can travel to different parts of the solar system in our search for life elsewhere than Earth.
Daniel Gregory, Assistant Professor,Department of Earth Sciences, University of Toronto
* * * * *
The paper: “A robust agnostic molecular biosignature based on machine learning”
Authors:
H. James Cleaves II
- Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC
- Department of Chemistry, Howard University, Washington, DC
- Blue Marble Space Institute for Science, Seattle, WA
Robert M. Hazen
Anirudh Prabhu
George D. Cody
- Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC
Michael L. Wong
- Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC
- NHFP Sagan Fellow, NASA Hubble Fellowship Program, Space Telescope Science Institute, Baltimore, MD
Grethe Hystad
- Mathematics and Statistics, Purdue University Northwest, Hammond, IN
Sophia Economon
- Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, MD
* * * * *
Photos
A carbon-rich black chert from Western Australia
Photo: https://bit.ly/3YWbZ4Z
One of the most tantalizing applications of the new method is the resolution of a decades-old debate regarding the origins of organic molecules in the 3.5-billion-year-old Apex Chert from the wilds of Western Australia.
This enigmatic black rock contains small quantities of carbon-rich residues–just enough to turn the chert a lustrous black. Some scientists have long argued that this formation holds the earliest record of cellular life in the form of tiny spheres and filaments – shapes that mimic modern microbes.
Other researchers insist that the black residues formed from high-temperature processes that have nothing to do with life. Research now in progress will apply the new biosignature method to the Apex Chert, as well as many other similarly enigmatic ancient rocks from Greenland, South Africa, India, and China.
* * * * *
Trilobites
Despite being 400-500 million years old, carbonized trilobite exoskeletons similar to these were sampled and clearly distinguished as biotic using this new analytical method:
1) Metacanthina sp. from Morocco, Devonian Period (~400 million years old), 7 cm maximum dimension
Photo: https://bit.ly/3P20Qfr
2) Koneprussia sp. from Morocco, Devonian Period (~400 million years old), 3.5 cm across.
Photo: https://bit.ly/45zUUji
3) Olenoides sp., Utah, Cambrian Period (~500 million years old), 8 cm maximum dimension.
Photo: https://bit.ly/3OyYyTq
4) Apianurus rusti, New York, Ordovician Period (~450 million years old), 5 cm maximum dimension.
Photo: https://bit.ly/3OGWvwL
(photo credits: Hazen Collection, National Museum of Natural History, Washington DC)
* * * * *
About Carnegie Science https://carnegiescience.edu/about
* * * * *
Coverage highlights
BBC, United Kingdom (154,922,793)
Aliens: Scientists use AI to search for extra-terrestrial life
https://www.bbc.co.uk/newsround/66945738
Daily Mail, United Kingdom (92,595,767)Is this the key to finding life beyond Earth? Scientists develop an AI system that can detect aliens with 90% accuracyhttps://www.dailymail.co.uk/sciencetech/article-12557735/Is-key-finding-life-Earth-Scientists-develop-AI-detect-aliens-90-accuracy.html
Forbes, United States (76,850,849)New AI Could Help Find Life On Mars With 90% Accuracy, Say Scientists https://www.forbes.com/sites/jamiecartereurope/2023/09/25/new-ai-could-help-find-life-on-mars-with-90-accuracy-say-scientists/
The Independent, United Kingdom (56,584,707)Scientists develop simple test to help us find alien life https://www.independent.co.uk/space/alien-life-extraterrestrial-mars-ai-b2418321.html
Sunday Mirror, United Kingdom (44,468,983)Scientists find way to confirm life on other planets in extraordinary alien breakthrough https://www.mirror.co.uk/news/world-news/scientists-find-way-confirm-life-31026629
London Evening Standard, United Kingdom (17,674,624)Scientists develop method of identifying life on other planets https://www.standard.co.uk/tech/science/scientists-earth-dna-sam-mars-b1109368.html
The Guardian, United Kingdom (3,367,304)Scientists hail pioneering software in hunt for alien lifehttps://www.theguardian.com/science/2023/sep/25/scientists-hail-pioneering-software-in-hunt-for-alien-life
Agencia EFE, via Vanguardia, Colombia (2,637,854)
La IA se convierte en la clave para detectar vida extraterrestre https://www.vanguardia.com/mundo/la-ia-se-convierte-en-la-clave-para-detectar-vida-extraterrestre-XN7618505
El Mundo, Spain (31,866,879)
La inteligencia artificial está aprendiendo a encontrar vida extraterrestrehttps://www.elmundo.es/ciencia-y-salud/ciencia/2023/09/25/65117c2afdddff02bb8b45b3.html
GEO, France (6,996,627)L’IA pourrait aider les scientifiques à détecter la vie sur Mars, et ailleurs dans l’univershttps://www.geo.fr/environnement/astrobiologie-ia-pourrait-aider-scientifiques-detecter-vie-mars-et-ailleurs-dans-univers-analyse-echantillons-curiosity-216800
The Messenger, United States (3,343,993)
Scientists Design AI To Detect Alien Life with 90% Accuracyhttps://themessenger.com/tech/scientists-design-ai-to-detect-alien-life-with-90-accuracy
Earth.com, United States (2,360,268)Is there really life beyond Earth? AI will soon solve this mystery https://www.earth.com/news/is-there-really-life-beyond-earth-ai-will-soon-solve-this-mystery/
El Especial, United States (3,521)La Inteligencia Artificial permitirá establecer si existe vida extraterrestre https://elespecial.com/la-inteligencia-artificial-permitira-establecer-si-existe-vida-extraterrestre
La Razón, Spain (13,940,544)¿Existió la vida en Marte? Una IA tiene la clavehttps://www.larazon.es/ciencia/existio-vida-marte_202309256511d95398383a00019d055f.html
El Independiente, Spain (4,696,708)El revolucionario método con IA para detectar vida en otros planetas: “Es el santo grial de la astrobiología”https://www.elindependiente.com/futuro/inteligencia-artificial/2023/09/25/el-revolucionario-metodo-con-ia-para-detectar-vida-en-otros-planetas-es-el-santo-grial-de-la-astrobiologia/
Asahi Shimbun via Yahoo!ニュース, Japan (89,937,333)90％で生命の痕跡を発見できるアルゴリズム開発–分子パターンの微細な違いを検出Algorithm development that allows 90 % to find traces of life -detection of fine differences in molecular patternshttps://news.yahoo.co.jp/articles/8c02086c88e9d68b352a73008055a0ad3e9198fe
Infobae, Argentina (62,231,555)Este método basado en inteligencia artificial detecta vida en otros planetasThis method based on artificial intelligence detects life on other planets https://www.infobae.com/tecno/2023/09/28/este-metodo-basado-en-inteligencia-artificial-detecta-vida-en-otros-planetas/
The Week via Yahoo! News, United States (61,429,064)The week’s good news: Sept. 28, 2023https://news.yahoo.com/weeks-good-news-sept-28-190313752.html
The Indian Express, India (43,451,001)Scientists develop new artificial intelligence method to look for life on other planetshttps://indianexpress.com/article/technology/science/artificial-intelligence-method-life-other-planets-8961690/
Detik, Indonesia (33,399,426)Ilmuwan Kembangkan AI untuk Deteksi Tanda Kehidupan di Luar Angkasa Scientists Develop AI for Detection of Signs of Life in Spacehttps://www.detik.com/edu/detikpedia/d-6955380/ilmuwan-kembangkan-ai-untuk-deteksi-tanda-kehidupan-di-luar-angkasa
O Estado de S. Paulo, Brazil (14,483,556)Inteligência artificial ajuda na busca por vida extraterrestreArtificial Intelligence helps in the search for extraterrestrial life https://www.estadao.com.br/ciencia/inteligencia-artificial-ajuda-na-busca-por-vida-extraterrestre/
Europa Press, Spain (10,659,065)
La IA ayudará a saber pronto si hubo vida en Marte(AI will help to know soon if there was life on Mars) https://www.europapress.es/ciencia/misiones-espaciales/noticia-ia-ayudara-saber-pronto-si-hubo-vida-marte-20230926113344.html
ANSA, via Tiscali, Italy (2,872,249)L’IA cerca nelle molecole tracce di vita aliena(The IA is looking for traces of alien life in the molecules)https://innovazione.tiscali.it/news/articoli/l-ia-cerca-nelle-molecole-tracce-vita-aliena/ … (ANSA) – ROMA, 26 SET – Analizzare qualsiasi tipo di molecola in cerca di tracce di vita alinea: riesce a farlo con una precisione del 90% …
Agenzia Giornalistica Italia, Italy (3,242,178)C’è vita nello spazio? L’AI potrebbe rispondere con una precisione del 90%(Is there life in space? AI could respond with a precision of 90%)https://www.agi.it/scienza/news/2023-09-27/vita-spazio-ai-potrebbe-rispondere-23194446/
IndoAsian News Service via Zee News, India (36,248,291) AI May Help Find Life On Mars, Other Planets With 90% Accuracyhttps://zeenews.india.com/technology/ai-may-help-find-life-on-mars-other-planets-with-90-accuracy-2667634.html
Der Tagesspiegel, Germany (9,704,421)Astrobiologie : KI soll die Suche nach Aliens revolutionierenAstrobiology: AI is supposed to revolutionize the search for alienshttps://www.tagesspiegel.de/wissen/astrobiologie-ki-soll-die-suche-nach-aliens-revolutionieren-10529587.html
Salon, via Yahoo! News, United States (61,429,064)A simple algorithm could reveal the ghosts of alien life in Mars dirt, study findshttps://news.yahoo.com/simple-algorithm-could-reveal-ghosts-160101168.html
Euronews, France (22,774,660)Is there alien life on Mars? Scientists believe AI test could help unlock the Red Planet’s secrets
https://www.euronews.com/next/2023/09/26/is-there-life-on-mars-scientists-believe-ai-could-help-unlock-the-red-planets-secrets
Space.com, United States (15,588,323)New AI algorithm can detect signs of life with 90% accuracy. Scientists want to send it to Mars https://www.space.com/new-ai-method-detect-life-rock-samples-90-percent-accuracy
Astronomy Magazine, United StatesCan AI find life in the universe? Sophisticated technologies are making inroads on the search for life in the cosmoshttps://www.astronomy.com/science/can-ai-find-life-in-the-universe/
Frankfurter Rundschau, Germany (10,348,052)
Leben auf dem Mars: Künstliche Intelligenz soll es finden können – „Das ist eine große Sache“ Life on Mars: Artificial intelligence should be able to find it – “that’s a big deal”https://www.fr.de/wissen/astrobiologie-leben-auf-mars-suche-kuenstliche-intelligenz-finden-biotische-materie-zr-92543084.html
UK Press Association, United Kingdom via The Independent (56,584,707), Scientists develop method of identifying life on other planets
https://www.independent.co.uk/news/science/scientists-earth-george-mason-university-dna-sam-b2418314.html
Live Science, United States (10,524,315)Scientists created AI that could detect alien life — and they’re not entirely sure how it works https://www.livescience.com/space/extraterrestrial-life/scientists-created-ai-that-could-detect-alien-life-and-theyre-not-entirely-sure-how-it-works
La Razón, Spain (13,940,544) ¿Existió la vida en Marte? Una IA tiene la clave https://www.larazon.es/ciencia/existio-vida-marte_202309256511d95398383a00019d055f.html
De Standaard, Belgium (6,772,483)Is er leven op Pluto? AI speurt mee https://share.belga.press/news/0034163a-d673-4c30-8db0-147c72f72f53
Full coverage summary, click here
News release in full, click here