Leveraging AI, Experts Find Human’s Third Ancestor

CNN Indonesia

Monday, 20 Feb 2023 18:45 WIB

Jakarta, CNNIndonesia —

For expert succeeded in identifying the third ancestor of humans as a result of interbreeding between modern humans and their descendants Neanderthals-Denisovans. How to?

Launch Eureka Alert, they found it by utilizing artificial intelligence technology (artificial intelligence).

The experts came from the Institute of Evolutionary Biology (IBE), Centro Nacional de Analysis Genomico (CNAG-CRG), and the Institute of Genomics at the University of Tartu.

They use algorithms deep learning combined with statistical methods in his research. As a result, they found hominids that may be the result of interbreeding between modern humans and descendants of Neanderthals-Denisovans.

However, experts do not know for sure about this. Hominids themselves according to the Big Indonesian Dictionary (KBBI) mean a tribe that includes humans and human-like creatures that have become extinct.

Neanderthals and Denisovans are known to have lived at the same time as modern humans more than 40,000 years ago in Eurasia.

“About 80 thousand years ago, the Out of Africa event occurred, namely when the human population – which includes modern humans – began to leave the African Continent. They migrated to other continents and spread the human population to what it is today,” said Jaume Bertranpetit, an expert from IBE as quoted from IFL Science.

“We know from then to the present that modern humans interbred with Neanderthals on all continents except Africa, and with the Denisovans in Oceania and possibly in Southeast Asia. However, evidence of interbreeding with a third, extinct species has yet to be confirmed,” he added. .

The use of AI for this research is quite important. This is because analyzing the possibility of transitions from DNA to ancestral demographic populations is too complex to be carried out with conventional methods.

Deep learning is an algorithm that mimics the workings of the mammalian nervous system, with different and specialized artificial neurons that learns to detect patterns that are important for performing certain tasks in data.

This study is the first to use deep learning to explain human history. This allows the use of this technology to be applied in other fields such as biology, genome and evolution.

“We used this technology to create an algorithm to learn to predict human demographics using the genome obtained through hundreds of thousands of simulations,” said Oscar Lao, an expert from CNAG-CRG.

“Whenever we run a simulation, we are walking through a path in human history. Of all the simulations, deep learning allows us to observe what makes ancient puzzles fit together,” he added.

(lth)