Han Lanqing, Executive Secretary General of the Alliance: Establishment of the alliance promotes the clinical transformation of rare disease drugs and empowers the diagnosis and treatment of rare diseases
AI technology helps to improve the efficiency of AAV capsid protein screening
Han Lanqing introduced that the key to treating rare diseases is to find the pathogenic gene. With current research methods, it is like finding a needle in a haystack to locate these genes. However, by using artificial intelligence to process and learn from the data, experimental results can be predicted, and the efficiency of drug development can be improved. As one of the most commonly used virus vectors in gene therapy, AAV has attracted a large number of pharmaceutical companies in recent years and made progress in a variety of rare disease indications. According to incomplete statistics, six AAV gene therapy drugs have been approved for marketing worldwide. In 2019, the FDA approved Zolgensma for the treatment of Type 1 Spinal Muscular Atrophy (SMA), which was the first AAV-based gene therapy for a central nervous system disease. Currently, eight AAV drugs that treat rare diseases have been approved for clinical use in China, mainly focusing on neurology, ophthalmology, and blood diseases, with the majority still in the clinical exploration stage.
Gene therapy is a technology that uses genes to treat or prevent diseases, including gene supplementation and gene editing. In recent years, gene therapy technology has gradually matured. With its characteristic of "curing thoroughly at one time," it can replace the mutated pathogenic genes of patients with normal genes. Therefore, it has become a key research direction for innovative treatments of rare diseases. Rare diseases have become the best choice for pharmaceutical companies to develop gene therapy for two main reasons: first, more than 90% of rare diseases still lack effective treatments; second, over 80% of rare diseases are genetic diseases caused by single gene mutations, which provide an excellent platform for gene therapy.
Currently, around 8,000 diseases have been identified as rare diseases, with approximately 80% caused by genetic mutations, mostly single-gene mutations. Gene therapy can target the cause of the disease to the maximum extent possible, making it more precise and effective. Therefore, AI can play a prominent role in the development of drugs for rare diseases. The key bottleneck of gene therapy is how to safely and effectively deliver the target gene into the body. Improving AAV production efficiency, developing new serotypes of AAV viruses that have better tissue specificity and higher infectivity, are of great significance for reducing AAV dosage, enhancing safety, and lowering treatment costs in gene therapy for rare diseases.
In this process, the modification of AAV capsid proteins is of utmost importance. "The most widely used method for modifying capsid proteins is targeted evolution assisted by artificial intelligence," Han Lanqing said. Recently published research showed that the technology team at Dyno Therapeutics, a biotech company in the United States, obtained 110,689 candidate mutants with capsid packaging ability from 28 amino acid fragments of AAV2 wild-type sequences through library design assisted by standard machine learning, far more than the number produced by rational design and random mutation libraries. This technology uses machine learning algorithms to analyze and model the screened data in the library and then generates new library sequences using machine learning models to improve the precision of library design, thereby improving the success rate of targeted evolution screening. "AI+ algorithms in the field of new drugs have been extended from machine learning to deep learning. Deep learning algorithms have solved the challenges faced by standard machine learning algorithms. In the process of drug discovery, they can be widely used for drug activity prediction, target discovery, etc. Using AI-guided deep learning technology to screen AAV capsid proteins for gene therapy delivery, hundreds of thousands of candidate mutants can be screened from one billion genes, increasing efficiency by several orders of magnitude. This is expected to become the first AI-based scenario application in AAV gene therapy delivery in China," Han Lanqing said.
Han Lanqing stated that traditional targeted evolution methods require the library to be synthesized and packaged into viruses before being injected into animals for screening. This process often needs to be repeated 3-6 times before a small amount of candidate vectors can be obtained. In contrast, AI-guided screening only needs to generate new AAV mutant sequences with target characteristics directly based on the set parameters after obtaining the first biological experimental screening and validation data. This technology overcomes the limitations of AAV mutant screening by DNA library synthesis and experimental techniques. Once this technology method is successfully established, it can quickly screen a large number of gene therapy vectors that can be used for clinical drug development in a short period, meeting the needs of most pharmaceutical companies in the market. Currently, this AI-assisted screening technology has been used by over 30 domestic pharmaceutical companies. Han Lanqing predicts that by 2025, it is expected that 5-10 rare disease drugs will enter clinical trials.
Currently, most rare disease drugs are very expensive, with a gene therapy drug for SMA in the US costing over 2 million dollars. Many rare diseases in China have no available treatments, and even if there are treatments available, they are expensive and have not yet been included in medical insurance reimbursement. If this screening method achieves expected results by efficiently screening AAV capsid proteins with target characteristics, it will not only provide treatments for genetic diseases but also significantly reduce the development costs and drug prices of gene therapy drugs, benefiting more patients with rare diseases.
It is reported that the "AI-assisted AAV capsid protein screening technology for gene therapy delivery vectors" developed by Han Lanqing's technology team participated in the 2022 National Subversive Technology Innovation Competition sponsored by the Ministry of Science and Technology. 2851 projects from across the country participated in the competition, and through multiple rounds of screening, 348 projects (with 14 projects from Guangdong Province) entered the national competition and are expected to enter the Ministry of Science and Technology's list of subversive technology candidates.
Figure 1. National Subversive Technology Innovation Competition
Establishing an alliance to promote the clinical transformation of rare disease drugs and empower rare disease diagnosis and treatment
Han Lanqing told reporters that currently, less than 10% of the 8,000 known rare diseases in the world have treatment plans. However, in China, there are only 55 drugs registered for rare disease indications, covering only 31 rare diseases listed in the "First Batch of Rare Diseases Catalog." Only 27 rare diseases are included in the national medical insurance directory. Being diagnosed with a rare disease is unfortunate, and having no treatment options is even more hopeless. Even if there are treatments available, they are often considered "expensive drugs" that only a small number of companies can afford to develop. If patients cannot afford these treatments, it will lead to the inability of companies to recover R&D costs, leading to less focus on developing treatments for rare diseases. Therefore, treating rare diseases is a circular question within the life and health industry chain.
Solving these problems requires the collective efforts of the entire industry chain to significantly reduce development costs, improve the quality of drugs/treatments, and expand drug production scale. This will enable the MAH (Marketing Authorization Holder) responsible for new drug research and development to complete preclinical evaluation, clinical trials, and market launch of their gene therapy drugs more efficiently and effectively, requiring less time and investment. This will also ensure that patients with rare diseases can access affordable gene therapy drugs earlier.
At the same time, in the field of rare diseases, there is generally a lack of research investment, low diagnosis rates, and a lack of effective treatment methods. Often, the few available treatments are not covered by the medical insurance system, leaving the vast majority of patients facing the dilemma of "no available treatment." Rare diseases are not rare, and this type of disease cannot be ignored in the history of public health and social and economic development. According to the 2018 Chinese Rare Disease Investigation Report, not only do ordinary people lack knowledge of rare diseases, but 33.3% of doctors explicitly stated that they also lack knowledge of rare diseases. It can be imagined that this proportion is even higher in primary medical institutions, which leads to a common misdiagnosis of rare diseases. To address this issue, in October 2021, the Zhuhai-Tsinghua Institute for Innovative Life Sciences, led by Han Lanqing, joined forces with universities, research institutions, hospitals, and pharmaceutical companies in Guangzhou and the Southern China region to establish the Guangzhou Rare Disease Gene Therapy Industry-University-Research Medical Technology Innovation Alliance (referred to as the Guangzhou Rare Disease Gene Therapy Alliance) to focus on researching genetic treatments for rare diseases.
Figure 2. The First Greater Bay Area Rare Disease Gene Therapy Summit Forum
The establishment of the alliance aims to bring together enterprises, research institutions, universities, and hospitals related to rare disease gene therapy to form a complementary cooperation and an integrated research and clinical collaboration network for rare disease gene therapy. This effort will accelerate the structural adjustment and optimization of China's rare disease gene therapy industry, with the goal of achieving early detection, early diagnosis, treatability, and manageability of rare diseases.
Technological innovation often stems from the intersection of disciplines and close collaboration between industry institutions. The establishment of the alliance can not only expand the existing research data management system through clinical feedback on rare disease cases but also fully leverage the advantages of each unit to produce a clustering effect. This will create a complete technical research system from clinical case resources-mutation site prediction-animal model construction-preclinical experiments of gene therapy-clinical gene therapy research.
The completion of the Human Genome Project has propelled the development of genome sequencing and other technologies, which are constantly applied to rare disease research and further promote the development of rare disease research. Second-generation genome sequencing technology, bioinformatics analysis, and other technologies can help reveal the pathogenesis of rare diseases and discover and identify pathogenic genes. However, 80% of rare diseases are caused by genetic variations. After conducting second-generation sequencing and bioinformatics analysis, the data will suggest a series of mutations that may be pathogenic. Researchers find it difficult to accurately determine which mutation sites ultimately lead to the onset of the disease among hundreds or even thousands of mutations. It usually requires conducting a large number of point mutation experiments on mice or rats, which often do not show phenotypes, resulting in significant waste of resources and time.
The alliance not only includes traditional industry chain-related units in rare disease gene therapy research, but also incorporates emerging strategic industries such as big data and AI to promote industrial structural optimization and upgrade and enhance the core competitiveness of the industry. For example, new mutation sites in clinical cases can combine with AI technology for more accurate pathogenicity predictions. Pathogenicity predictions combined with interspecific comparisons in bioinformatics can guide us to construct animal models more similar to human diseases. AI technology can also screen gene therapy vectors with more diversity, targeting, and lower immunogenicity. By using optimized vectors to import disease models more similar to humans for therapeutic research, we can better achieve clinical translation.
During industrial clustering, new patents will arise in various areas such as rare disease data information processing and optimization, rare disease animal model construction, AI-optimized AAV vectors, gene therapy drug delivery, etc. Bringing together various enterprises, research institutions, hospitals, and other institutions that were previously fighting alone and integrating them into an efficient rare disease gene therapy research and development technology chain, to establish core technology standards.
After more than a year of effort, the joint development of the Guangdong-Hong Kong-Macao Greater Bay Area Institute for Innovative Life Sciences and Sibiono BioTech has led to the creation of the Rare Disease Data Center (RDDC, https://rddc.tsinghua-gd.org/). It integrates epidemiological data, drug development profiles, disease-related gene maps, gene mutation sites, disease-related experimental mouse models, and other data and information for confirmed rare diseases worldwide. The RDDC takes diseases, genes, and animal models as the main line and displays multiple dimensions of related data and information based on data collection, collation, and reconstruction for rare diseases.
In addition, RDDC fully utilizes existing publicly available genetic big data resources and has developed genetic disease-related gene diagnostic tools based on AI and bioinformatics technology. The first publicly available AI tools include Patho Predict, which predicts the pathogenicity of mutations, and RNA Splicer, which predicts mutation splicing. This is also the first time these AI tools have been made available in the form of a free web page.
In 2023, the Guangzhou Rare Disease Gene Therapy Alliance looks forward to recruiting more fresh blood and invites outstanding companies and industry elites to join us in supporting the diagnosis and treatment of rare diseases.
Figure 3. Han Lanqing, Executive Secretary-General
of the Guangzhou Rare Disease Gene Therapy Alliance
Han Lanqing is the Director of the Artificial Intelligence Innovation Center at the Zhuhai-Tsinghua Institute for Innovative Life Sciences and the founder of Sibiono BioTech. He is also a major national talent, an excellent entrepreneur in Innovation China, an overseas high-level returnee, a member of the first Immunology and Cell Therapy Committee of the Chinese Association for Laboratory Animal Science, a member of the second Guangdong Provincial Experimental Animal Standardization Committee, the Vice Chairman of the Seventh Council of the Guangdong Provincial Experimental Animal Society, a leading talent in innovation and entrepreneurship in Guangzhou, and an artificial intelligence expert in the expert database of Guangzhou Industrial and Information Technology Commission, among others. He has more than 30 core intellectual property rights currently owned or applied for, has led/participated in over 10 government-level scientific and technological projects at the national, provincial, and municipal levels and has published research results in authoritative industry journals such as J Allergy Clin Immunol, EBioMedicine, and Gastroenterology.
"Journey with Rarity"
The "Journey with Rarity" interview program is jointly produced by the Guangzhou Rare Disease Gene Therapy Alliance and the Southern Medical Network. The program invites authoritative experts, researchers, entrepreneurs, and other professionals in the field of rare diseases to conduct interviews on the latest research, technological progress, clinical prevention and treatment, and other aspects of rare diseases. It provides industry professionals and researchers with popular science knowledge and cutting-edge information related to rare diseases, and helps to promote progress in the diagnosis and treatment of rare diseases.