Fig 3. Median pAUCs in relation to the number of samples in the different modalities scenario.

Each line corresponds to a specific method. The x axis represents the number of samples per group used in the simulations, and y axis the median pAUCs. Circles suggest that the observed differences from the best method were not statistically significant at level ≤ 0.05. A) Only genes which exhibited the same behaviour across datasets were taken into account. The integrative analysis yielded better results regardless the employed method. CP and omicsNPC^Fisher were the best method in most cases. B) All genes were considered in this analysis. Again, omicsNPC^Fisher and CP achieved better performance than other methods in most cases; the Benjamini method was also statistically indistinguishable from the best performing method when the sample size was 6 or 10.