His work to capture and display the dynamic small scale behavior of invisible biomolecules was awarded Zhuoran Qiao as the first Chen Institute and Zhuoran Qiao. Science Awards from Al Accelerated Research. The award recognizes innovative young researchers who apply techniques to artificial intelligence to help the research community solve important problems and accelerate their work.
“We are excited to partner with Chen Institute to launch this new award initiative,” said Yury V. Suleymanov. Science. “Our winner Zhuoran Qiao shows outstanding achievements in this field. His research introduces transformative approaches to deciphering and reprogramming molecular biology using artificial intelligence-driven structural basic models. This helps AI overcome the limitations of traditional methods and shows how it can pave methods for new opportunities in molecular design and therapeutic development.”
Machine learning and molecular structure
Interactions between biomolecules such as proteins and small molecules are key to supporting the fundamental processes of life. Identifying these interactions on an unprecedented scale can be useful in developing new drugs, both in other applications, but it requires deciphering the three-dimensional structure of these interactions. This requires a zoom-in snapshot of the molecular compartment.
Traditional methods for determining molecular structures such as X-ray crystal structures and cryogenic electron microscopy are powerful but slow. It may take several months of work in the lab to generate important molecular images.
Recently, AI-driven protein structure prediction tools have made strong advances in this regard. They can predict the three-dimensional structure of a protein from its amino acid sequence. However, these new tools represent the beginning of our journey to creating a full-scale “computer microscope” for molecular biology,” said Qiao, founder scientist at San Francisco-based artificial intelligence startup Chi Discovery. It's important to see things on the scale of the biomolecules in the system, not just 100, but thousands of atoms, and various conformations, he said.
February 2024 Nature Machine IntelligenceQiao and his colleagues have advanced the work of predicting AI-driven protein structures to date by developing a new generation machine learning approach to create a clearer view of two important activities of two important activities. Protein ligand interactions and the landscape in which these interactions occur
When a ligand, a molecule or ion that binds to a central atom or molecule, performs the binding task, it affects the structure it binds to.
“If you want to develop a new drug, you need to model the interactions of biomolecules really accurately,” says Qiao. “You need to get the structure right and understand how strongly the two proteins or small molecules interact. This is the first thing you need to know if a drug will succeed.”
It's a very complicated task, he added. “To show how molecules move in real life is like navigating a very complex maze with thousands of dimensions.”
Introducing NeuralPlexer
The Qiao team, developed to visualize these interactions, is called NeuralPlexer. We consider that biomolecules are highly dynamic and require numerous snapshots to fully capture the behavior. Therefore, this tool gradually improves the fine details of the structure that are produced, starting with an initial sketch of the entire molecular complex. The process, Qiao said, allows researchers to “quickly get a full picture of molecular interactions with the details of atomism.”
Qiao and his colleagues used NeuralPlexer to perform tasks such as predicting the formation of special binding sites that would not exist unless spurred ligand binding. They demonstrated a powerful tool to identify new drug binding pockets, among other tasks.
“Comparing this approach to a traditional approach offers what a high-throughput method does in six months per day,” he said.
Qiao was motivated by this work from the early days of the field on the basis of his perception that scientists understand the theoretical frameworks of many systems they study in computational chemistry, but many of the related problems are in practice not able to be calculated.
“It's a great honor for me to win an award,” Qiao said. “It's a great recognition of my chosen research path. It's also deeply humble, as it reminds me to continue doing impactful work, such as mentoring others to become interested in computational chemistry. New technologies demonstrate the real translation potential for computational research to develop better drugs and healthcare.”
Qiao is pleased to be part of how molecular modeling is changing despite still having a lot to do. He wants to address some of the next steps in Chay Discovery.
“We were thrilled to receive such impressive applications from around the world, spanning many different science disciplines,” said Chrissy Luo, co-founder of Chen Institute. “As AI is fundamentally accelerating global scientific discoveries, we are pleased to work with AAAS to introduce three incredible young researchers using these powerful new technologies to expand the frontier of human knowledge.”
Finalist
Finalists for the award include Adityania, Assistant Professor Nanyan at Nanyan Technology University in Singapore. The second finalist of the award is Aliséreu Bad, a fellow of the Flanders Marine Institute. Her work aims to improve our understanding of the role of oceans in the global carbon cycle, using machine learning to monitor coastal networks, particularly along coastal coasts.
