A reverse-engineering approach to dissect post-translational modulators of transcription factor’s activity from transcriptional data

Table 1 Kinase subfamilies predicted by DMI to modulate the 14 TFs used for validation

TFs	Subfamily Predictions
CDX2	PIM, MAPK [db], DMPK, CDC2/CDKX [db], SYK/ZAP-70, Lammer, VRK
E2F	MAPK [db], CSF-1/PDGF receptor, CaMK
ELK1	CSF-1/PDGF receptor (0.001), MAPK [db]
ETS1	CaMK [db], HIPK, MAPK [db]
GATA1	CaMK, HIPK, MAPKK [24], GCN2, MAPK [db]
GATA2	CaMK, MAPK [db], DMPK, SRC [25]
MYC	CaMK [26], CSF-1/PDGF receptor [27], MAPK [db], HIPK, GCN2, SRC [28]
SMAD3	DMPK [db], CSF-1/PDGF receptor [27], MAPK [db], PIM, Lammer, CaMK [db]
SMAD4	CaMK [db], DMPK, MAPK [db], PIM, HIPK [29], GCN2, SRC [db]
STAT1	CaMK [db], BUB1, STE20
STAT3	CSF-1/PDGF receptor [db], DMPK, SYK/ZAP-70
STAT6	EGF receptor [30], Fibroblast growth factor receptor [31], I-kappa-B kinase, CSF-1/PDGF receptor [32], MAPKKK [33], JAK [db], AXL/UFO
TCF4	CaMK [34], DMPK, MAPK [db], PIM, HIPK
TP53	CSF-1/PDGF receptor [35], Lammer [db], MAPK [db], DMPK [db]

In bold, subfamilies correctly identified by DMI as confirmed either by literature or by a phospho-interactome database [db] (Material and Methods). Kinase subfamilies are sorted according to the p-value of their enrichment score and results have been cut with a p-value threshold of 0.01

ISSN: 1471-2105