Screening data analysis

Screening data was analyzed using an in-house developed, open-source, Phyton-based high throughput screen (HTS) analysis pipeline (all documentation and code available under: https://github.com/elkeschaper/hts). Net-FRET data was calculated [21] and subjected to various quality control checkpoints. Initially, a heat map of individual plates was plotted to examine temperature-induced gradients or dispensing errors. Subsequently, z’-Factor and strictly standardized mean difference (SSMD) scores [24,25] were calculated to report the robustness of the screens. Additionally, Net-FRET values were plotted to check for row or column effects as well as assessing correlation of duplicates or triplicates. After assessing quality of each individual plate, candidate genes were selected with the following cut-off criteria: SSMD of < -4 and > 4 or a p-value (t-test) of below 10⁻¹⁵. After the secondary screen, same cut-off criteria were used with the additional requirement at least 2 out of 3 individual siRNAs targeting the same transcript passing the threshold. Graphs were generated with GraphPad Prism.