Who Was the Hybrid Teenage Girl?
In 2019, archaeologists from Russia found a suspicious part of a skeleton in a cave around Siberia. After protein analysis confirmed it was a hominin bone, researchers from the Max Planck Institute found that the bone fragment belonged to a thirteen-year-old girl.
Apparently, she lived 50,000 years ago and was a hybrid of modern humans. Her mum was a Neanderthal, and her dad was a Denisovan, so the researchers named her Denny (or, more formally, Denisova 11).
What’s unusual is that up until now, there wasn’t actual proof that Neanderthals and Denisovans have crossed paths, interacted and had mutual descendants.
Although initially from the same group, they separated 390,000 years ago and evolved into distinct groups. What’s more, the existence of Denisovans was only discovered in 2010, so for researchers to find an actual offspring between the two species is more than impressive.
The finding of a first-generation Neanderthal–Denisovan offspring among the small number of archaic specimens sequenced to date suggests that mixing between Late Pleistocene hominin groups was common when they met.
Slon, V., Mafessoni, F., Vernot, B. et al. The genome of the offspring of a Neanderthal mother and a Denisovan father.
Nature 561, 113–116 (2018). https://doi.org/10.1038/s41586-018-0455-x
How Can AI Help Scientific Findings?
Introducing new technology into scientific research means we can approach what’s already known with a new perspective, but also analyze new findings with advanced techniques.
That’s how results of a research conducted by M. Mondal from Estonia and J. Bertranpetit and O. Lao from Spain confirmed the third introgression in all Asian and Oceanian populations from an archaic population. They used deep learning algorithms to build a demographic model, using a statistical technique called Bayesian Inference.
By combining deep learning algorithms and statistical methods, investigators from the Institute of Evolutionary Biology (IBE), the Centro Nacional de Análisis Genómico (CNAG-CRG) of the Centre for Genomic Regulation (CRG) and the University of Tartu have identified, in the genome of Asiatic individuals, the footprint of a new hominid who cross-bred with its ancestors tens of thousands of years ago.
As they suggest, Denisova 11 was not an isolated case but rather a hint of a more general introgression process.
This study was the first to use deep learning technologies to inspect human evolution. In the future, it’s expected to become more common in fields such as biology, genomics and evolution.
“Besides faster image processing, using AI in paleontology can help establish research standards,” said Congyu Yu from the American Museum of Natural History.
How Can Deep Learning Explain Human History?
You may have heard that we’re all genetically related at a time depth of up to 300 thousand years ago and share a common African root.
Around 80,000 years ago, a part of the human population migrated from Africa to other continents – also known as the Out of Africa event.
Now, even though we have evidence that Neanderthals and Denisovans have cross- bred, the authors of this study also insinuate that these two species couldn’t produce fertile descendants.
The existence of a third ancestor was only a theory until now, but deep learning has been able to identify that this newly found species interbred with modern humans tens of thousands of years ago.
Photo Illustration: Freepik
Even though deep learning has made it possible to make the transition from DNA to the demographics of ancestral populations, the investigators experienced some serious challenges.
That’s because these demographic models were so much more complex than anything else considered to date, so there were no adequate statistical tools to analyze them.
By the words of the principal investigator Òscar Lao, deep learning was able to imitate the exact way the nervous system of mammals works, with different artificial neurons that specialize and learn to detect, in data, important patterns for performing a given task.
“Whenever we run a simulation, we are traveling along a possible path in the history of humankind. Of all simulations, deep learning allows us to observe what makes the ancestral puzzle fit together”, he concluded.