Next PostThe Periodic Jet Precession Is Successfully Captured, and the Latest Research Findings on M87 Black Hole Are Published in Nature
Modern pharmacy has a history spanning over a hundred years. More diseases and pathological conditions can be treated or alleviated by pharmacological intervention. However, the successful development of new drugs is time-consuming, laborious, and expensive. Therefore, how to efficiently design safe and effective drugs has become an urgent problem for scientists.
On October 12, Nature Computational Science published a paper titled A Universal Programmable Gaussian Boson Sampler for Drug Discovery, the latest results of the collaborative research by Zhejiang Lab (ZJ Lab), Imperial College London and the University of Science and Technology of China. The journal also wrote an op-ed opinion review for the research achievement, titled Programmability Empowering Quantum Boson Sampling. This paper presents a programmable Gaussian Boson Sampling (GBS) photonic quantum processor developed by the research team, and the research progress in designing two important drugs that undergo successful molecular docking and RNA folding prediction. These results are further proof of the huge potential of quantum computing in industrial applications such as drug research and development.
Advantages of Computing
In fact, computing technologies for drug research and development have been quite mature. Large-scale computing and complex simulation are indispensable for drug research and development that ranges from molecular modeling, drug screening to efficacy prediction. However, when computing and simulating large-scale complex biomolecular systems, especially when the quantum effect is involved, the difficulty and complexity of computation increase exponentially. Traditional computing approaches often try to solve problems by finding approximations at the expense of accuracy.
A Schematic Diagram of Quantum Computing and Molecular Docking
"With regard to complex systems that cannot be effectively handled by traditional computing approaches, quantum computing has inherent advantages in directly and efficiently simulating quantum systems in the real physical world." YU Shang, the first author and co-corresponding author of the paper, as well as the postdoc at ZJ Lab, said: "It means that quantum computing can capture details of the interaction between drugs and biomolecules and simulate interactions between molecules more realistically to help us find potential drug candidates more quickly, predict their biological activity more accurately, and optimize their efficacy and safety more effectively."
"Fully tapping into the potential of quantum computing represented by the Gaussian Boson sampler, accelerating the launch of new drugs, and reducing R&D costs will bring more and better treatment options to patients," said Raj Patel, an important research collaborator and a Future Leaders Fellow from Imperial College London.
Photonic Quantum "Abacus"
The abacus is a great invention in ancient China and also one of the earliest calculating tools for humans. After vicissitudes of life, it holds out great vitality. Manipulating beads on an abacus can be regarded as a process by which computations are objectified and recompiled. Basic operations are represented as rules by which beads move up and down and advance and retreat, and multiple operations can be combined to perform complex computations.
A Schematic Diagram of Universal Programmable Photonic Quantum Processor
This photonic quantum processor is also named Abacus. "Based on the 'time window' encoding mode, we have endowed the photonic quantum processor with universal programmability that can encode arbitrary structural 'graph' parameters into the processor via an electro-optical modulator to execute operations," YU Shang introduced. Thanks to the efforts of ZHONG Zhipeng, a senior engineer at ZJ Lab, the working process of Abacus is fully programmed, which consumes less computational resources and is highly scalable. This has the same effect as traditional abacus.
The research team conducted several rounds of experiments on Abacus to verify its capabilities. Researchers programmed graph structures based on specific molecules into Abacus, and used GBS to find maximal fully-connected subgraphs from these graph-structured data. Abacus can draw out the best docking pose of a drug with viral protein from identified maximal fully-connected subgraphs. In addition, during communication with researchers from the Georgia Institute of Technology, the team found that Abacus can also be used to predict RNA foldability. The sampling results reveal that Abacus can find maximal fully-connected subgraphs within an "all-weighted stem-and-leaf plot" representing RNA sequences very efficiently, and thus obtain optimal predictions about RNA-folded structures. These results indicate that Abacus excels in searching best docking poses for drugs and viral proteins and predicting RNA-folded structures.
Molecular Docking Results Obtained with Abacus
RNA Folding Predictions Results Obtained with Abacus
A Long Way to Go
At present, the industrial use of quantum computing is still facing multiple challenges, including yet-to-be-overcome key core technologies such as quantum error correction, weak foundation for special-purpose algorithm models, and shortage of quantum device fabrication technologies. This achievement fully proves quantum computing's potential for drug research and development, and the research team is confident that quantum computing will further drive changes and progress in related industrial applications.
"Adhering to the concept of open science, we have gathered global scientific research forces, including the research team of Fellow Ian Walmsley at Imperial College London, the research team of Academician GUO Guangcan at the University of Science and Technology of China, and researchers from universities in Canada and the United States. We believe that with our joint efforts, new computational tools and algorithms like Abacus will spring up. This will promote interdisciplinary collaboration among physics, computing science, biology and pharmacy, allowing for a new generation of industrial applications such as drug research and development, and bring more efficient and innovative treatment solutions to people suffering from diseases around the world." DONG Ying, a research fellow at ZJ Lab, said: "We will make more efforts on technical practicability to push quantum computing towards practical and commercial application."
Dr. YU Shang and Senior Engineer ZHONG Zhipeng from ZJ Lab, and Dr. FANG Yuhua from the University of Manitoba in Canada contributed equally to this paper as the first authors, and ZJ Lab is the affiliation of the first author and corresponding author.
Dr. YU Shang, Research Fellow Raj B. Patel from Imperial College London, as well as Dr. WANG Yitao, Prof. TANG Jianshun, and Prof. LI Chuanfeng from the University of Science and Technology of China are co-corresponding authors.
The main collaborators also include Research Fellow DONG Ying, Dr. XU Liang, and Engineer LI Qingpeng at ZJ Lab; Prof. Geoffrey K Tranmer at the University of Manitoba; Dr. LIU Wei at the University of Science and Technology of China; LI Zhenghao and Ewan Mer, PhD students at Imperial College London; as well as Dr. TANG Mengyi and Prof. Sung Ha Kang at the Georgia Institute of Technology.
This work was supported by the National Natural Science Foundation of China, China Postdoctoral Science Foundation, EPSRC QCS Hub, UKRI, and EU Horizon, etc.