Artificial intelligence to identify chemical signs of life on other worlds

2023-09-27 16:45:05

There is a lot of scientific and popular interest in looking for telltale signs of life on other worlds, but how will we recognize them when we catch them? Scientists have developed a system based on artificial intelligence that, at least in the tests carried out, achieves more than 90% accuracy when discovering signs of life.

This advance is the work of Robert Hazen’s team, from the Carnegie Institute and George Mason University, both institutions in the United States.

Hazen argues that with the new system, human civilization’s ability to recognize chemical signals of life on other stars will increase significantly. And it opens a path towards the routine use of intelligent sensors on board space vehicles to provide them with insight far superior to what they have today when identifying such chemical signals that reveal the presence of life.

Since the 1950s, it has been known that, given the right conditions, mixing simple chemicals can form some of the more complex molecules necessary for life, such as amino acids. Since then, many components required for life, such as the nucleotides needed to make DNA, have been detected in space. But how do we know if these components are of biological origin or if they are formed through some abiotic process over time? Without being able to elucidate this origin, we will not be able to know if we have detected traces of life.

From an evolutionary point of view, life is not something easy to maintain, and that is why there are certain paths that work and others that do not. The research carried out by Hazen and his colleagues is not based exclusively on the identification of a compound, but on determining whether or not the origin is biological by analyzing the compound in relation to the context of the sample in which it is present.

Hazen and his colleagues used pyrolysis gas chromatography methods combined with mass spectrometry, used by NASA on Mars (in the Viking probe mission and in the Curiosity robotic rover mission).

In this photograph taken on Mars by the Curiosity robotic rover, part of it and the place it was exploring appears. (Photo: NASA JPL / Caltech / MSSS)

Using these methods, Hazen’s team analyzed 134 varied carbon-rich samples from living cells, samples degraded by the passage of time, geologically processed fossil fuels, carbon-rich meteorites, and laboratory-synthesized organic compounds and mixtures. Of them, 59 were of biological (biotic) origin, such as a grain of rice, a human hair, crude oil, etc. The rest (75) were of non-biological (abiotic) origin, such as laboratory-synthesized compounds, including amino acids, or carbon-rich meteorite samples.

The samples were first heated in an oxygen-free environment, causing them to decompose (a process known as pyrolysis).

The treated samples were then analyzed in a device that separates the mixture into its components and then identifies them.

Using a set of machine learning methods (a form of artificial intelligence) and data from each abiotic or biotic sample, the artificial intelligence system was trained, with the result that it achieved the ability to infer the abiotic or biotic nature of the sample with greater than 90 percent accuracy.

The results of this research were made public at the last Goldschmidt Geochemistry Congress, held in France. (Source: NCYT from Amazings)