Figure 4.

Library coverage as a function of FDR and library equivalents (ε) screened. Screening simulations were run for a 100k-member library with a 1.5% hit rate. Screening aliquot sizes ranged from 1−10 library equivalents (100 simulations). Library coverage curves for each cumulative replicate class (k ≥ 1–5, gray) are based on a Poisson cumulative distribution function. FDR (%) was plotted as a function of ε for each cumulative replicate class (k ≥ 1–4, teal). Coverage (bold trace) constrained by FDR (<5%, gray horizontal) is calculated for each simulation iteration and the average is plotted (standard deviation markers are smaller than the data points). Simulation results are shown for average droplet occupancy λ_drop = 1.