Fig. 7: A PECAn-powered RNAi-based screen identifies a list of cell competition genes.
a Schematic showing how target genes were identified. Initial hits were identified as those which were upregulated relative to the wildtype in wing discs heterozygous mutant for one of either of two loss-of-function alleles for RpS3 and in homozygous Mahjong mutant wing discs. These initial hits were then refined by comparing them against genes upregulated in a third prospective loser condition, wing discs mildly overexpressing Nrf2. Of these 121 genes, we obtained 91 RNAi lines from the VDRC KK collection targeting 87 genes. b List of gene hits identified as candidate modulators of cell competition in the screen. MiWO wing discs expressing an RNAi were compared against control MiWO discs for both percent pouch coverage and percent cell death coverage at RpS3+/− patch borders. Hits were identified as those which yielded a statistically significant effect, as determined via a Wilcoxon–Mann–Whitney U-test with FDR p-adjustment, on either parameter. c Post-screen validation of a subset of gene hits, showing replication of screen results for RNAi’s targeting ab, Gclc, Gdh, or Vajk2, which exacerbate RpS3+/− cell competition and Zormin, which alleviates RpS3+/− cell competition. Statistics for the percent patch coverage reflect two-sided student’s t-test with an un-pooled Hedges’ g effect size metric, and statistics for the percent caspase coverage of the border reflect two-sided Wilcoxon–Mann–Whitney U-tests with a Cliff’s δ effect size metric. Box and whisker plot denotes minimum, first quartile, median, third quartile, and maximum, whereas red diamond and whiskers denote mean and 95% CI. Biologically independent samples per condition are as follows: Abrupt: ncontrol = 15, nRNAi = 15; GCLC: ncontrol = 11, nRNAi = 12; Gdh: ncontrol = 8, nRNAi = 5; Vajk2: ncontrol = 9, nRNAi = 7; Zormin: ncontrol = 14, nRNAi = 10 (replicate 1), ncontrol = 10, nRNAi = 10 (replicate 2). Source data are provided as a Source Data file.
