Fig. 5
From: MBE: model-based enrichment estimation and prediction for differential sequencing data
Simulated negative selection prediction results. Comparison of wLER and MBE predictive accuracy for simulated negative selection using the 100-unit NNs on the a 21-mer insertion (\(4.6 \times 10^7\) short reads), b avGFP mutagenesis (\(4.6 \times 10^5\) long reads), and c AAV recombination (\(4.6 \times 10^5\) long reads) datasets. Dot size represents the fraction of test sequences with the lowest ground truth fitness used to compute Spearman correlation. A 0.2-fraction of lowest test sequences corresponds to the sequences in the bottom 20% of all unique test set sequences after ranking by ground truth fitness. To produce each curve, we computed Spearman correlation for every fraction of lowest test sequences between 0.01 and 1.0, inclusive, in 0.01 increments for each method, to obtain a sweeping continuum in the plot, although the legend shows dot sizes only in increments of 0.2 to reduce clutter. The dashed black line represents the equal performance of the two approaches