Pooled FLIM screening to efficiently elucidate cell signalling

Objectives

1. In a pooled screening concept, we will use (existing) viral CRISPR gene knockout libraries to effectively knock down one single gene per cell in a large population of cells. This results in a mix of ~ 21,000 knockdown cells representing all genes (except for the lethal ones). We have shown that using the new FLIM instrumentation we can perform dynamic screens at ~150,000 cells in parallel on a single coverslip, and we developed technology to analyse signals and pick out individual hits (cells with an aberrant phenotype) within a single afternoon. Thus, a 100-fold covering of the entire genome should be feasible within 1-2 weeks already, and this may be further optimized.  

2. Given the vast amount of data generated using this screen we will use innovate machine learning algorithms to classify knockouts in a step-wise and logical sequence c) Focusing initially on the cAMP response signalling cascade; we will rank the machine learning parameters according to their statistical weightings and delineate specific pathways using conventional cell biological methods.

Host Academic Institution: Netherlands Cancer Institute