Skip to main content
Fig. 2 | Genome Biology

Fig. 2

From: gscreend: modelling asymmetric count ratios in CRISPR screens to decrease experiment size and improve phenotype detection

Fig. 2

Computational simulation explains how cell splitting causes asymmetry of before/after ratios. a Schematic representation of the simulation. After generation of an initial gRNA abundance distribution, different functions (blue) are applied to model transduction, cell growth, cell splitting, and sequencing. The simulation outputs the gRNA counts obtained by sequencing the plasmid library as well as the cell pools at time points T0 and T1 (green, R1 and R2 are technical replicates). The simulation depends on a set of user-defined parameters (yellow, see Table 1). b–e Simulation results for different values of cell splitting coverage Ccells and cell doubling time τ, while other parameters remain fixed. Only gRNAs without fitness effects are shown. b gRNA abundance at T1 compared to T0 for simulation with Ccells of 100 and 800. gRNAs with large observed fold changes are colored (LFC <−1 in pink, LFC >+1 in green). c, e Fraction of gRNAs with LFC <−1 (left) and LFC >+1 (right) for Ccells ranging from 100 to 1500 (c) and τ ranging from 20 to 90 h (e). Mean over 5 simulations is depicted. d MAGeCK-RRA precision-recall curves on data simulated using different values for Ccells (100, 400, and 1500). The recall at 95% precision is indicated. f Schematic representation of cell splitting during the proliferation phase of screen, which consists of multiple rounds of exponential growth and random sampling. g Count distribution of gRNAs targeting non-essential genes at T08, T15, and T18 of the screen in HCT116 cells [2]. gRNAs were ranked according to their abundance and the resulting ranks normalized to [0;1] (library fraction, x-axis). On the y-axis, the counts per gRNA are shown

Back to article page