From Data to Cure: Leveraging Artificial Intelligence and Big Data Analytics in Accelerating Disease Research and Treatment Development

Abstract

The emergence of novel genotyping, big data, data sharing, open-source algorithms, and powerful computing resources has facilitated the integration of artificial intelligence (AI) and big data to elucidate disease mechanisms, find implements for drug design and development, and enhance precision medicine. The parallel explosion of published studies containing large quantities of biomolecular data with accompanying clinical or pathology information has indeed benefited disease studies. However, similar to the biases that plagued many of the datasets used for the development of commercial AI algorithms, diagnosing radiological images, biological oddities or data imbalances could mislead the outcome of AI and big data studies. There is a high need of standardization and heterogeneous curation practices in the input data for AI and big data prometheus. Therefore, a systematic overview of the key technological advances, challenges, and emerging solutions for data mining is warranted for benchmarking and motivating further development of AI and big-data-guided disease research.

As the window of opportunity for the rapid advance of artificial intelligence (AI) solutions to emergent challenges in drug discovery and global health narrows, there is an immediate need to harness big data, AI, and shared research infrastructure to optimize the performance of AI systems on small data sets in biomedicine. The coordination of such systems is a massive and complex engineering challenge that must safeguard researcher freedom while preventing misuse and enhancing collaboration. Effective governance of AI in the life sciences and biomedicine requires specifying requirements and mechanisms to monitor the evolution of each constructed system with respect to agreed-upon principles.