The fourth scientific discovery paradigm for precision medicine and healthcare
The scientific discovery paradigm (SDP) provides a mature and routine framework for asking scientific questions, developing methods or strategies to answer such questions, and also includes ways to explain the experimental results or the observed data. In the last two decades, the SDP in life sciences has shifted fast, especially with progression of the human genome project. A paradigm shift, sometimes also called a 'scientific revolution', occurs when the existing paradigm cannot efficiently solve the challenges faced and a new paradigm is needed to deal with the challenges. For example, in bioinformatics, a small paradigm shift we refer to here as the bioinformatics scientific research model (SRM), emerged with accumulation of DNA sequencing data. Since then, a batch of new genes has been discovered by pattern identification with models trained using known gene structure patterns. Well-known bioinformatics tools and databases including CLUSTAL W,1 MEGA,2 PDB3etc., were developed within the bioinformatics SRM. Traditional experimental paradigms can only discover new genes one by one through time-consuming and labor-intensive methods. Complex biological systems, however, often function by interactions between many genes, proteins, or other components via pathways, modules, or networks. Bioinformatics has contributed to acceleration in life sciences by fast, efficient, high throughput, and computational methods, enabling investigation of biological and medical problems at systemic levels. The microarray, yeast two-hybrid assay, and evolutionary modeling promoted the paradigm shifting to systems biology, which aimed to reconstruct the interaction or synergistic network to explain emergence properties in a system. Systems biological tools such as gene ontology,4 KEGG5 and Cytoscape,6etc., were then developed and widely used. But for clinical translation, genome functional discoveries cannot be applied directly to treatment of patients because of heterogeneities among diseases and patients. Cell-line or animal-model based biological findings need to be validated with patient samples before clinical applications. Therefore, translational and precision medicine SRMs have been proposed to integrate genotypic and phenotypic information for personalized prediction and treatment of diseases.7, 8
Although paradigms in life sciences have shifted frequently in the past 20 years, data accumulation is always the driving force for scientific revolution. In the future, data will remain one of the most essential parts for successful scientific paradigm shifts; however, the quality, quantity, and diversity of biomedical data will pose key challenges for our future precision medicine and healthcare.
