Emerging strategies to bridge the gap between pharmacogenomic research and its clinical implementation

摘要

Cells are transfected with a diversity library of expression plasmids covering many (ideally all possible) amino acid substitutions for a given pharmacogene of interest and a protein-specific selection assay is applied. Deep sequencing of the transfected cells before and after selection, and comparative analysis of the obtained data can provide enrichment scores for each variant, thus enabling the massively parallelized experimental characterization of thousands of variations. Importantly, this data provides a powerful resource for the training and optimization of computational algorithms, including but not restricted to random forests and deep neural networks, which in turn can be applied to the integrated analysis of a patient’s entire pharmacogenome.