The key concepts of precision medicine are prevention and treatment strategies that take individual molecular profile and clinical information into account. Single-cell next-generation sequencing technologies (NGS), liquid biopsy for circulating tumor DNA (ctDNA) analysis, microbiome, radiomics, and other types of high-throughput assays have exploded in popularity in recent years thanks to their ability to produce an enormous volume of data quickly and at a relatively low cost. The emergence of these big data has pushed forward the goal of precision medicine; however, from the entire scope of capture and utilization, including the electronic health records (EHR) and big data analysis on high-throughput biomarkers, to the ultimate goal of clinical usage based on a patient’s genome, there are still many more challenges ahead.
In recent years, almost all top biomedical journals have published major findings using the advanced data science technologies including complex statistical modeling, machine learning, and Artificial Intelligence (AI). The challenges of how to interpret the results to patients and the true clinical utilities remain.
　　The application of AI and Machin Learning (ML) to medicine will lead to evolutions in clinical practice. US FDA Commissioner Scott Gottlieb, MD, has forecasted on Twitter that AI and ML hold enormous promise for the future of medicine. Since 2017, several products that leverage AI-based technology for clinical application have been approved by the US FDA. We may expect the US FDA to approve more and more clinical applications of AI and ML. However, the reproducibility of recently reported promising results may be in question, absent the formulation and application of principles for training dataset integrity and the assessment of the generalizability on real word data.
In this presentation, I will introduce several initiatives on precision medicine in the United States, including Vanderbilt University’s BioVU initiative, and the U.S.’s All of Us (previously known as the Precision Medicine Initiative). I will also offer some perspectives on the changing landscape for precision medicine, including the road map for choosing between statistical modeling and machine learning, the concept of treating the unstructured text as quantitative data, physicians’ mind resetting about the explosive growth in information technology, machine learning, and the AI revolution. In addition, I will talk about future developments in medicine, including how to design and conduct pivotal trials, pragmatic trials, and real world evidence studies. These areas present great opportunities for medical researchers to strengthen their role in the precision medicine. I’ll finish up with some potential pathways of seamlessly integrating molecular, cellular and genomic data with clinical, physiological, behavioral and environmental parameters in the precision medicine era.