Scientists have long been fascinated by convergent evolution
– that mind-blowing moment when totally unrelated species come up with the same winning feature. Think of 🦇 bats and 🐋 toothed whales, both mastering echolocation like natural sonar, even though they split from each other millions of years ago.
Now, a Chinese mainland research team from the Institute of Zoology of the Chinese Academy of Sciences, led by Zou Zhengting, has decoded a key part of this evolutionary game using AI. They developed a computational framework called ACEP, powered by a pre-trained protein language model. This AI tool acts like a life scientist’s ultimate cheat code, understanding the hidden structural and functional patterns in amino acid sequences.
“A protein language model can understand the deeper structural and functional characteristics and patterns behind amino acid sequences,” Zou explained. By analyzing high-order protein features, ACEP revealed why different organisms hit upon the same solution—echolocation—when adapting to similar environments.
Their findings, recently published in the Proceedings of the National Academy of Sciences, not only crack a chapter of nature’s playbook but also show AI’s superpower in tackling complex biological puzzles. 🤖🔬
Looking ahead, the team hopes to expand AI’s role in evolutionary biology and unlock even more secrets of life’s adaptation and innovation.
Reference(s):
Chinese scientists use AI protein model to decode convergent evolution
cgtn.com
