Table 4.

cis-regulatory motifs (and corresponding transcription factors) consistently associated with up-regulated gene sets from all three experiments in Table 1

Motif	Transcription factor	Combined P value (πc)	AHB soldierP value	Alarm pheromone P value	Old bee P value
All differentially expressed genes
CACGCG	Hairy (Dmel)	1.35E-05	0.00012	0.00016	0.007
Trl	Trl (Dmel)	4.49E-05	0.00058	0.004	0.0088
WHWWWWWW	unknown	8.69E-05	0.00035	0.0063	0.0119
V_AHR_Q5	AHR (Hs)	0.00015	0.00035	0.0138	0.011
V_ETF_Q6	ETF (Hs)	0.00035	0.0066	0.0037	0.027
Deaf1	Deaf1 (Dmel)	0.00048	0.0149	0.028	0.00067
MGAAD	Hsf (Dmel)	0.00074	0.0107	9.21E-05	0.043
As above, minus those common to two or more gene lists
CACGCG	Hairy (Dmel)	0.00043	0.00025	0.0085	0.032
MGAAD	Hsf (Dmel)	0.00058	0.0099	0.00027	0.038
V_CDPCR1_01	CDP (Hs)	0.00062	0.038	0.0085	0.0028
Ap	ap (Dmel)	0.0009	0.0019	0.0073	0.05
Trl	Trl (Dmel)	0.00097	0.0017	0.04	0.02
Klf4	Klf4 (Hs)	0.0012	0.042	4.5E-05	0.026
CACNNG	unknown	0.0017	0.0064	0.045	0.031
Deaf1	Deaf1 (Dmel)	0.0018	0.0098	0.072	0.00055

Shown are the P values of association for each experiment (Fisher exact test, see SI Text). The significance of each motif association across all experiments is shown in Column 3 (π_c), based on the combined measure P_c from Columns 4–6. The corresponding empirical P value π_c(P_c^(o)) and its FDR for multiple hypothesis correction was calculated (see SI Text). Only motifs with FDR <0.001 are reported. The bottom half of the table shows eight most-significant motifs discovered from the same procedure if the gene sets were purged of any genes that belonged to two or more gene sets. Only soldiers were analyzed because their brain gene expression profiles were most similar to those of alarm-pheromone-exposed and old bees. Transcription factor motifs from: Dmel, Drosophila melanogaster; Hs, Homo sapiens.