Chinese scientists use AI protein language model to uncover secrets of evolution

A team of Chinese researchers have identified how different species independently evolve similar traits, revealing the potential of artificial intelligence (AI) in advancing evolutionary biology, according to Xinhua News Agency, a partner of TV BRICS.

The study has shed new light on how distantly related species such as bats and toothed whales have independently developed echolocation. Using an AI protein language model, the scientists have identified a key mechanism explaining how different organisms can evolve similar traits when adapting to comparable environments, a process known as convergent evolution.

Led by Zou Zhengting from the Institute of Zoology of the Chinese Academy of Sciences, the team developed a computational framework called ACEP, which employs a pre-trained protein language model. This technology enables the analysis of deeper structural and functional patterns within amino acid sequences, revealing the crucial role of high-order protein features in adaptive convergence.

According to the researchers, the study highlights not only the complex laws of life’s evolution but also the growing potential of AI in solving intricate biological problems. Their findings could pave the way for broader applications of AI in understanding evolution and biodiversity.

Photo: RyanKing999 / iStock