One of the greatest enigmas that has remained since the beginning of the space age has been that of the existence of life on other worlds.
At present, there are various strategies that science has implemented in its quest to find extraterrestrial life: Tracking of radio emissions from the cosmos, search for chemical traits in the atmospheric composition that planets that are close to another star may have, and analysis of samples of celestial bodies, among others. These are some of the methods put in place to find that vestige of extraterrestrial life that we have longed for for decades.
However, it is necessary to consider that, in the case of alien life, given the differences with terrestrial life forms, this cannot be detected under methods that use certain terrestrial substances as biosignals as criteria.
To solve this dilemma, a team of scientists led by Tomohiro Mochizuki, from the Tokyo Institute of Technology in Japan, carried out a study focused on the development of a machine learning technique, which works by examining complex organic mixtures through spectrometry of masses, complemented with an ultra-high resolution that allows them to be classified as biological or abiological.
This is how the researchers put their technique to the test by examining a wide variety of complex organic mixtures from meteorites, which have organic compounds that were generated abiologically.
They also included laboratory microorganisms, unprocessed natural crude oil collected from oil wells that is then used to be refined and converted into gasoline, and abiological samples created in the laboratory.
The researchers then subjected all the raw data from these complex mixtures to a computerized machine learning algorithm. In the end, a classification of the samples was obtained in which it was possible to determine with 95% precision which were of living or non-living origin.
You can get more information at noticiasdelaciencia.com.