Table 1.

Area under the ROC curve in synthetic data

Differentially spliced exons	Noise level σ	PECA-SI	RMA-LM	PLIER-LM	RMA-SI	PLIER-SI	RMA-MIDAS	PLIER-MIDAS	FIRMA
1	0.7	0.99	0.99	0.98	0.99	0.94	0.99	0.95	0.99
2	0.7	0.99	0.98	0.96	0.98	0.94	0.98	0.94	0.98
3	0.7	0.97	0.94	0.91	0.93	0.93	0.94	0.93	0.93
4	0.7	0.94	0.87	0.83	0.86	0.90	0.90	0.90	0.86
5	0.7	0.92	0.83	0.79	0.82	0.88	0.87	0.88	0.81
1	1.0	0.94	0.91	0.89	0.97	0.92	0.90	0.85	0.94
2	1.0	0.94	0.91	0.88	0.93	0.89	0.91	0.86	0.91
3	1.0	0.90	0.86	0.82	0.86	0.82	0.87	0.81	0.84
4	1.0	0.88	0.83	0.80	0.84	0.80	0.85	0.79	0.82
5	1.0	0.74	0.68	0.68	0.69	0.69	0.72	0.67	0.66

The synthetic data were generated according to Equation 8 at two different noise levels, σ = 0.7 or σ = 1. The first column indicates the number of synthetic differential splicing events generated within a gene. At each combination of the noise level and the number of differentially spliced exons, 1,000 genes were investigated. In each case, the probe-level PECA-SI procedure (see Equation 6) was compared to the standard SI procedures RMA-SI, PLIER-SI, RMA-MIDAS and PLIER-MIDAS, the two-way ANOVA-based approaches RMA-LM and PLIER-LM, and the FIRMA algorithm. The largest area under the curve (AUC) value across the methods is indicated in bold.