Loading...
Loading...
2 mins read
A pioneering study using artificial intelligence models has revealed how the brain has evolved over 320 million years, identifying similarities between birds and mammals. These findings not only offer new insights into intelligence and cognition but also open up new possibilities for the study and treatment of neurodegenerative diseases such as Alzheimer’s and Parkinson’s.
The human brain, one of nature’s most fascinating and complex structures, has its roots in an evolutionary past that stretches back hundreds of millions of years. A recent study led by Prof. Stein Aerts and his team at the VIB-KU Leuven Center for Brain & Disease Research in Belgium has used deep learning models powered by artificial intelligence to explore how the brain has evolved over the past 320 million years.
Published in the journal Science, this research not only sheds light on the mechanisms that have shaped intelligence and cognition across species, but also offers promising tools for better understanding neurological diseases.
The researchers analyzed the regulatory codes that control gene activity in brain cells of humans, mice, and chickens. These codes, which act like switches to turn genes on or off, determine how brain cells function. Using deep learning techniques, the team trained artificial intelligence models to decode these genetic instructions, discovering that some types of neurons have remained virtually unchanged for over 320 million years—a remarkable example of evolutionary conservation.
On the other hand, other cell types have undergone significant changes, adapting to the specific needs of each species.
A particularly significant finding was that certain neurons in birds share regulatory codes with deep-layer neurons in the mammalian neocortex—the brain region associated with advanced cognitive functions. This discovery suggests that intelligence and cognition may have developed from shared evolutionary mechanisms, despite the anatomical differences between bird and mammal brains.
The use of artificial intelligence has not only enabled the comparison of genomes across species to identify both conserved and divergent regulatory codes, but has also opened new opportunities in the study of neurological disorders. In previous research, Aerts’ team demonstrated that the regulatory codes of certain cellular states in melanoma are conserved between mammals and zebrafish, allowing them to identify genetic variants associated with the disease.
Now, their models are being applied to the study of neurodegenerative diseases such as Parkinson’s and Alzheimer’s, by analyzing the relationship between genetic variants and cognitive traits. By predicting which genetic mutations affect brain function, researchers can develop more effective strategies for diagnosing and treating neurodegenerative conditions.
Moreover, these models can analyze genomes to detect the presence or absence of specific cell types in any species, offering a powerful tool for studying both evolution and medicine.
This advancement in genetic analysis and cellular regulation not only deepens our understanding of brain evolution but also holds significant medical potential. As scientists expand these models to a broader range of species and human pathological states, new insights are expected into how the brain functions and how it has been shaped by genetics throughout evolutionary history.