Skip to main content
. 2020 Apr 30;15(4):e0232271. doi: 10.1371/journal.pone.0232271

Fig 1. Performance comparison of 13 DE analysis methods for SEQC benchmark data and simulation data based on KIRC parameters.

Fig 1

Three and five samples were used in simulation tests. (A) SEQC spike-in data test, 5 sample condition, 0.25, 1.5, 2 fold changes are introduced to DE genes. (B) KIRC based simulation. 1.5 or larger and 1.3 or larger fold changes are introduced to 0.27% of 10,000 genes for 3 and 5 samples, respectively. (C) Same condition as B but fixed fold changes, 0.625, 1.15, and 1.3 were introduced to DE genes. Dispersions were also lowered to SEQC level (22.5 times).