The last decade has seen the introduction of precision medicine, which emphasizes the search for and use of molecular signatures to aid disease diagnosis and treatment, and translational medicine, which promotes the translation of clinical research into clinical application. Precision medicine has been made possible by the landmark results of important global collaborative projects such as the human genome, human proteome, human epigenome, human microbiome, cancer genetic atlas and many others. The rapid progress in translational medicine has also been accelerated by the rapid development of data-driven artificial intelligence. The area that has benefited most directly is the development of clinical multi-omics, which builds mathematical models driven by medical data and clinical knowledge for a given clinical problem by integrating phenotypic data such as electronic medical records, physical examination records and molecular features from multi-omics.

Such a strategy has dual applied and theoretical implications. On the one hand, through the establishment of such models, validated molecular signatures and clinical features can be screened, thus laying the data basis for the establishment of new clinical tests and diagnostics, thus facilitating the medical translation of concomitant diagnoses and clinical decisions. On the other hand, the screening of the molecular signatures and clinical features can be further scrutinized and interpreted to find mechanistic support for disease mechanisms and drug development, and to facilitate the development of the pharmaceutical industry.

Externally, Shenzhou Medical has established extensive and strong links with international and domestic precision medicine institutions, clinical medical organisations and renowned multinational pharmaceutical companies. Not only does it have an office next to the Broad Institute in Cambridge, Boston, USA, but it also has a commercial partnership with Congenica®, a spin-off of the Wellcome Trust Sanger Institute near Cambridge University, UK, and a strategic partnership with Sentieon® in Silicon Valley, California, USA.

Internally, the translational precision medicine data science departments in Beijing and Shanghai work closely with the precision medicine operations departments across the country and in Singapore to develop clinical multi-omics research and related data integration and analysis platforms with domestic and international medical institutions, ultimately resulting in a business model that benefits patients, doctors, hospitals and companies alike.

*  The Translational Precision Medicine Data Science department is small in size but well structured.

*  The leaders of the department have been trained in the US for over a decade in industry and academia in the US and China, in addition to their systematic and specialist cross-disciplinary PhD training. They have a deep understanding of not only biomedical data science, but also scientific computing, information processing, molecular mechanisms and drug design in life, clinical and pharmaceutical sciences, and have a keen awareness of the frontiers and advances in these fields.

*  The team includes not only experienced data analysts in histology, but also data scientists who are skilled in data-driven learning algorithms and knowledge scientists who are familiar with molecular pathways of disease and drug knowledge as well as medical statistical methods.

*  For each translational precision medicine project, professionals from different backgrounds work together to effectively communicate with collaborating medical experts and researchers to rapidly advance the project.

*  In addition, the team members have developed and built a commercial-grade, high-volume multi-omics data integration and analysis platform for population genomics based on specific collaborative projects.

*  Data analysis in precision and translational medicine involves two layers.

*  The first is the analysis of multi-omics data, such as point mutation data, gene expression data, methylation data, copy number variation data, flora data, protein profile data, metabolomics data, etc., but also histology data including single cells, to extract meaningful molecular features and integrate them into a structured clinical record.

*  This huge data matrix is then reprocessed and data mined based on the data itself and known clinical knowledge to construct mathematical models for a given clinical problem and to find the meaningful features.

*  The features are first screened for specific clinical problems and can therefore be rapidly industrialized for the next step of large-scale clinical trials and regulatory approval. At the same time, these screened features can be used to guide the design of disease mechanisms or drugs for specific clinical problems.

*  To accomplish this two-tier data analysis, a commercial, high-volume multi-omics data platform is indispensable. Building such a commercial platform can greatly enhance the efficient translation of clinical precision medicine.