Fig. 3: Nearly complete genome-wide cell fitness gene assignment by IntAC.
A Expression level of fitness genes (FPKM + 0.01) assigned by genome-wide arrayed RNAi screens27, v.1 CRISPR screens, v.2 IntAC CRISPR screens, or v.2 CRISPR screens without anti-CRISPR. Genes arranged from low to high maximum likelihood estimation Z-scores and binned in groups of 1000 genes. Since only expressed genes can be fitness genes, v.2 IntAC screens lead to superior fitness gene assignment, and leaving out anti-CRISPR severely diminishes fitness gene identification. B Detection of essential genes across components of essential cell structures—spliceosome, ribosome, and proteasome—in v.1 and v.2 screens. While these genes were enriched in v.1 screens, a substantial portion was missed. In contrast, the majority of the core components were detected in the v.2 IntAC screens, illustrating the improved sensitivity of the IntAC system for detecting essential cellular processes. The color of each component reflects its gene Z-score. C Precision-recall curves for fitness gene assignment in v.1 screens (green), v.2 screens (purple), or genome-wide arrayed RNAi screens (gray) using components in B as true-positives and non-expressed genes as false-positives. v.2 IntAC screens exhibited a recall rate of 90–95% for essential genes, compared to 35–60% in v.1. This marked improvement in recall demonstrates that the IntAC system yields more comprehensive and accurate identification of fitness genes across the genome. Genome-wide RNAi screen data is from ref. 27. Elements created in BioRender. Viswanatha, R. (2025) https://BioRender.com/orvd36w.
