BGI: AI medical leader leads a new journey of intelligent life

AI is reshaping the landscape of the medical industry. Data shows that from 2019 to 2023, the size of China’s AI medical market climbed from 2.7 billion yuan to 8.8 billion yuan, with a compound annual growth rate of 34%. In the emerging market of AI medical, companies represented by BGI are rooted in gene sequencing, actively promote the in-depth integration of AI technology with genetic testing, disease diagnosis and other businesses, and have established market-leading technological advantages.

On February 7, 2025, BGI organized an investor relations event to elaborate on the company’s layout in the AI field through a telephone exchange meeting, and share its cutting-edge exploration and achievements in the integration of artificial intelligence and life sciences.

BGI actively builds the world’s leading multi-omics database, providing rich data resources for the training and optimization of AI models. Through the development of GeneT, a multimodal large model for gene testing, ChatGeneT genomic consultation platform, and an intelligent disease prevention and control system 13311i, it has occupied the leading position of AI medical care.

Model technology continues to innovate and AI results continue to be exported

Since the advent of the big model, the medical industry has been regarded as the industry most likely to have rich scenarios implemented. AI technology empowers the medical industry. On the one hand, it reduces the originally high medical costs and reduces the pressure on medical insurance and patients; on the other hand, with the support of the large model, medical resources between regions are balanced, allowing more people to enjoy high-quality medical resources that they would not otherwise have enjoyed.

At present, in many fields such as genetic testing, BGI’s large model and other products have successfully embarked on the stage of clinical application. By improving the efficiency of pathogenic site selection, interpretation costs have been greatly reduced. BGI’s large model technology has two main application directions: one is to accelerate bioinformatics analysis and research based on basic model frameworks such as Transformer, and the other is to use open source large language models for clinical applications.

In September 2024, BGI proposed the generative biological intelligence paradigm GBI ALL (Generative Bio-Intelligent), which aims to use the power of intelligent technology to accurately gain insight into and efficiently apply life data to accelerate the intelligent process of genetic testing and health management. Under this paradigm, systems such as GeneT(Genetic Transformer), a multimodal large model for genetic testing for comprehensive intelligent interpretation, and ChatGeneT, a genomics consultation platform for the public.

GeneT, a multimodal large model of genetic testing, has used more than one million levels of high-quality data to build a tens of billions of high-quality token. Combined with the experience of interpretation experts, it has achieved accurate interpretation of the whole genome data. The ChatGeneT Genome Consulting Platform aims to shorten the distance between professional knowledge and the public. Through an intelligent consultation system, it provides users with convenient and accurate genome interpretation services. Ordinary users can upload their own genetic testing reports to the ChatGeneT platform. Learn about the genetic risks of related diseases and prevention suggestions through dialogue.

In addition, BGI has also conducted in-depth cooperation with Peking Union Medical College Hospital on large models of rare diseases, and is committed to re-analyzing undiagnosed patients through AI technology to explore potential genetic mutations. At the same time, combining basic medical research and laboratory verification to ensure the accuracy of prediction results can accelerate drug research. The large model has a large number of parameters and can combine multiple genetic methods and polygenic disease detection, which is expected to bring significant benefits to the diagnosis of polygenic diseases.

These results point out the direction for the integration of genetic technology and AI technology, and are also important milestones in the exploration of related tracks.

Relying on data assets to consolidate AI medical competitive advantages

Although many companies have participated in the AI Medical track, in comparison, BGI has the advantage of having the ability to produce fully domestically produced genetic data. The genetic testing process can independently and reliably generate genetic data at controllable costs. In addition, it has a great first-mover advantage in data accumulation and data analysis. Based on the data and algorithm research and development capabilities accumulated in the past two decades, the company can better combine professional knowledge on the basis of general models, especially in integrating multiple health management inspection data and conducting automated and intelligent analysis, which greatly improves the detection efficiency. It also ensures extremely high accuracy and stability through actual verification of tens of thousands of samples.

At the same time, BGI also mentioned at the exchange meeting that the diversity of large-scale real-world data sources and the comprehensiveness of the real case database are the key to improving the accuracy of the model to clinical level availability. It passes tens of thousands of samples. Practical verification ensures extremely high accuracy and stability, and applies artificial intelligence technologies such as prompt word engineering, fine-tuning enhanced learning and enhanced indexing technology in algorithm development. By combining enhanced indexing (RAG) technology, the knowledge base and case base accumulated by BGI can be used more fully and effectively, greatly improving the generalization ability and accuracy of the model.

In terms of external cooperation, BGI has built a training data set of millions of cases through cooperation with top hospitals, including public data and commercial cooperation data. The high cost of data acquisition and preprocessing constitutes a high technical barrier. Through informatization construction and data alignment, BGI ensures data structure, reliable and stable sources, and alignment, thereby outputting high-quality data. This kind of data processing not only meets current needs, but also considers users ‘needs for repeated use of data in the future, thereby achieving standardization and facilitating acquisition, storage and mining. It can be seen that BGI has significant competitive advantages in the field of AI medical, allowing it to maintain a leading position in the fierce market competition.

In addition, BGI has also emerged its strong application potential for deep integration of AI technology into the medical field in multiple key business segments.

In the past few years, BGI has had a good layout and business growth in its early tumor prevention and control business. It has broken through the prevention and control model of early screening of different cancer species through the most suitable combination of technologies from different groups, and is committed to using genomes, multi-dimensional data such as proteome and metabolome promote the development and application of more effective early tumor screening products. In the research and development of multi-cancer early screening technology, AI mainly plays a role through two key links: first, it uses multi-omics technology to discover extremely small amounts of cancer signals; second, it can accurately determine the location of the signal source, namely traceability and positioning. By combining public large models with vertical domain data training, a more economical and effective model can be established to reduce parameter fitting to obtain better results and improve the specificity and accuracy of the model.

At the same time, BGI has bravely taken the lead and used AI technology to make medical protection of personal health more accurate. At the exchange meeting, investors asked: For personalized diagnosis and treatment, is it possible for everyone to have their own genetic coding sequence so that they can provide guidance on their own diagnosis and personalized medication? BGI’s reply is that the whole genome information is the basis for personalized diagnosis and treatment. It can provide information such as drug response, disease risk prediction, disease prevention and aging-related indications. In terms of personalized medication, BGI has mature business applications, such as it is used for personalized treatment of cancer patients, and is gradually expanding to efficacy prediction and prognosis judgment. It can be seen that AI technology is expected to inject new vitality into the development of the entire medical industry by analyzing patient test data and push the medical industry to new heights.

Launched the ALL plan to shift from production-based services to intelligent services

In BGI’s view, the reduction in sequencing costs caused by the independent control of tools in the field of genomics is accelerating the arrival of the era of everyone’s genome. Faced with massive genomic data, how to efficiently and accurately extract the required genetic information to understand the disease process is a major problem that needs to be solved urgently. BGI effectively combines past genomic data with AI algorithms, which is conducive to precise health management for medical institutions and individuals, accelerates the development of precision medicine, and promotes the process of digital intelligence in life sciences.

At the exchange meeting, BGI mentioned that it will be committed to building a core AI intelligent medical system in the future to help transform and upgrade various business segments. In 2025, we will comprehensively shift the business model from production-oriented services to intelligent services around the core strategic goals of the ALL program, promote the automation and intelligence of front-end sample processing, and combine short-read-long and long-read-long sequencing technology platforms to further enhance database accumulation. Lay a solid foundation for future business transformation.

From a broader perspective, with the significant drop in sequencing costs and the continuous advancement of technology, the application prospects of high-depth whole genome data in the medical field are becoming increasingly broad. BGI’s future business model will also adapt to the times, promote it combined with localized characteristics, and create a diversified business model to adapt to the needs of different countries and regions. Through large-scale clinical verification, it will ensure the performance and data of products in the real world. The accuracy of the data will then outline the genetic blueprint, accurately anchor the health course, and lead the trend of AI medical care.

In addition, as an aging society intensifies, social and economic growth and the gravity of the continuous expansion of market space will promote the rapid development of the AI medical industry. As people’s health awareness becomes increasingly awakened and requirements for the quality of medical services continue to rise, their desire for AI medical products and services will continue to grow, especially in the fields of livelihood projects such as early cancer screening and elderly health management. Leading medical companies such as BGI rely on cutting-edge technology, massive data and excellent innovation capabilities to continuously explore solutions to medical problems and keenly capture market demand. It is expected to become a sharp blade to overcome medical problems and a strong new driving force to open up broad market space, leading the medical industry to a new stage of development.