Table 1.
Top 3 causal networks predicted by IPA34 explaining expression changes in the input dataset.
| Master regulator | Participating regulators | Depth | Predicted activation | Activation z-score | p value of overlap | Network bias-corrected p value | Target molecules in dataset |
|---|---|---|---|---|---|---|---|
| PPARγ | APP, NFkB (complex), PGR, PPARG | 2 | 0.707 | 5.84E−04 | 2.1 E−03 | APP, ATP1B1, BMI1, CD59, CTSL, CYP3A4, HSPA1A/HSPA1B, SLC1A2 | |
| KIF5B-RET | CHUK, ERBB2, ESR1, FGFR1, IKBKB, … | 3 | Inhibited | − 2.324 | 6.19 E−04 | 5.3 E−03 | APAF1, APP, ARHGAP1, ATP1B1, BMI1, CD59, CDC23, CSAD, CYP3A4, … |
| MUC1 | ADAM17, AKT1, CASP8, CASP9, CDKN1A, … | 3 | − 1.528 | 7.07 E−04 | 7.1 E−03 | APAF1, APP, ARHGAP1, ATP1B1, BMI1, CASP2, CD59, CDC23, CSAD, … |
The network bias-corrected p value of overlap (calculated using Fisher’s Exact Test, p value < 0.05) defines possible significant overlaps between the input genes and the known targets regulated by an upstream regulator.