Skip to main content
Fig. 3 | Genome Biology

Fig. 3

From: Response to "Neglecting normalization impact in semi-synthetic RNA-seq data simulation generates artificial false positives" and "Winsorization greatly reduces false positives by popular differential expression methods when analyzing human population samples"

Fig. 3

Comparison of DE methods on semi-synthetic data (with varying sample sizes) generated using the permutation-based strategy from GTEx data of heart left ventricle vs. atrial appendage under the data generation scheme 2. The FDR control (left panel) and power given the actual FDRs (right panel) for a range of per-condition sample sizes from 2 to 100, under FDR thresholds (i.e., claimed FDRs) of 10% (top panel) and 0.1% (bottom panel). The claimed FDRs, actual FDRs, and power were all calculated as the averages of 50 semi-synthetic datasets with the same per-condition sample size

Back to article page

Genome Biology

ISSN: 1474-760X

Contact us