Supplementary Figure 1. 208 Chip-Seq/Cetch-Seq

1.0

maximum TAF1 possible 0.8 fraction of all TSSs POL2

0.6 ARID4B SAP130

0.4

P300

Fraction overlapping with TSS regions TSS with overlapping Fraction 0.2 FOXA1 FOXA2 MAFF

0.0 0 20000 40000 60000 80000 ChIP-seq peaks called

Total human TSS regions = 44,488

Supplementary Figure 1. 208 ChIP-seq/CETCh-seq experiments plo9ed by number of peaks called in each experiment (x-axis) vs fracEon of peaks overlapping with any of 44,488 TSSs in the human genome (peaks +/- 3 kb of TSS). Selected individual factors are labelled. Solid line is linear regression through all points; do9ed lines represent number of total TSS regions and maximum possible fracEon of TSSs. Supplementary Figure 2. IDEAS segmentaEon of HepG2 genome. LeT: color key for all IDEAS states. Right: Pie chart indicaEng fracEon of HepG2 genome associated with each state. Supplementary Figure 3. Clustering of DNA associated factors based on chromaEn states recapitulaEng the assigned cluster with PC1 (63.50%), PC2(16.51%) and PC3(6.48%) variances explained. Supplementary Figure 4. Number of factors with called peaks within 2 kb of the TSSs of 57 liver speciﬁc genes with expression levels in HepG2 shown by bar color. TPM = Transcripts Per Million. A 5 B 5

4 4

3 3 log(10) TPM log(10) TPM 2 2

1 1

0 0 0 50 100 150 200 0 50 100 150 200 TFs bound at promoter TFs bound at promoter C 1.0 D 1.0

0.8 0.8

0.6 0.6

0.4

Rank percentile 0.4 Rank percentile 0.2 0.2

0.0 0.0

PAF1+ PAF1- ATF4+ HSF1+ HSF1- ATF4- A

Exact TF number match Random percentile ranks

Within 5% of TF number match POLR2AphosphoS2+POLR2AphosphoS2-

Supplementary Figure 5. A. Sca9erplot of all genes (black points), showing log10 TPM in HepG2 (y-axis) vs number of unique TFs with called peak +/- 2 kb of gene TSS (x-axis). Blue line indicates linear regression through black points. Red points represent 57 liver-specific genes; red line indicates linear regression through red points. B. same as A, but all genes expressed below 10 TPM are removed; blue line indicates linear regression through only these >10 TPM genes. C. DistribuEon of rank percenEles for expression of 57 liver-specific genes, compared to exactly matching number of TFs (leT box) and to within 5% of number of TFs (center box); random rank percenEle for comparison is shown in right box (Mann-Whitney p-value < 0.0001 for both exact and 5% match when tested against random). D. Rank percenEle of expression of all genes with specific TF’s presence compared to rank percenEle of equal number of random matched genes with within 5% of same number of TFs but without specific TF. TFs analyzed are PAF1 (M-W p<0.0001) , ATF4 (M-W p=0.0093), POLR2AphosphoS2 (M-W p<0.0001), and HSF1 (M-W p=0.0002). TPM = Transcripts Per Million. Boxes indicate quarEles, whiskers are drawn to 5-95% quarEles. Supplementary Figure 6. DistribuEon of regulatory regions by number of associated TFs (leT). DistribuEon of horizontally matched sites by IDEAS states (right). A

B C

Supplementary Figure 7. CumulaEve fracEon of called moEfs in our data compared to moEfs in databases as scored by TomTom similarity E-value in A) CISBP (build 1.02) Homo sapiens database, B) JASPAR 2016 vertebrate database, and C) JASPAR 2018 vertebrate database. 180

160

140

120

Annotation CTCF Bound 100 Enhancer−like Heterochromatin and Repressed Count 80 Promoter and Enhancer−like Promoter−like

concordance discordance no_match

Supplementary Figure 8. DistribuEon of TF moEfs by concordance (matching expected TF), discordance (matching diﬀerent TF), and no match in CIS-BP database. Stacked bar plots are colored by main TF groups from previous unsupervised clustering. A B

Supplementary Figure 9. A) DistribuEon of TF moEfs highly dissimilar to all moEfs in CIS-BP (y-axis) and their median oﬀset distance from the center of peaks (x-axis). B) Stacked distribuEon of highly dissimilar moEfs (no match; green) with similar (concordant; blue) and moEf called for secondary factor (discordant; red) and their median oﬀset distances from the peak center (x-axis). A Motif Status B Associated Factors

FOXP1 RREB1 THRA RARA ELF3 GATA4 FOXA3 ERF RXRB BCL6 TEAD1 ZNF219 FOXA2 RXRA MYBL2

s SOX13.iso1 r MLX Occupancy Fraction SOX5 0.8 PROX1 0.6 FOXA1 acto Cooccupancy Fraction ARID5B F 0.4 0.8 CUX1 0.2 0.6 MIXL1 Forkhead Factors SIX4 0 0.4 GATAD2A FOXO1 0.2 ZKSCAN8 Primary motif (+) ZGPAT Primary motif (−) ZSCAN9 NFIC Associated RERE FOXK1 HMG20A NCOA2 ETV4 THRB SP5 ERF MLX SIX4 NFIA NFIC ELF3 ETV4 BCL6 SOX5 CUX1 THRB THRA RXRA RERE RXRB RARA MIXL1 GATA4 ZGPAT SOX13 NR2F2 FOSL2 TEAD1 CEBPA MYBL2 HNF4A PROX1 RREB1 HNF4G NCOA2

SP5 ZNF219 ARID5B ZSCAN9 HMG20A GATAD2A CEBPA 1 ZKSCAN8 NFIA NR2F2 0.8 HNF4A HNF4G 0.6 FOSL2 0.4 ●

0.2 X only O No motif Self only F 0 Both motif

Associated Factors Associated Factors

C FOXP1 D

FOXP1

FOXO1

FOXO1 s r FOXK1

FOXK1 acto F

FOXA1 Normalized Fraction Normalized Fraction 0.5 FOXA1 0.5 0.4 0.4

Forkhead Factors 0.3 0.3

0.2 Forkhead 0.2 FOXA2 0.1 FOXA2 0.1

FOXA3 FOXA3 T A A P A2 A4 X5 X1 P T X13 SP5 SP5 O O O ERF ERF MLX MLX SIX4 SIX4 A NFIA NFIA NFIC NFIC ELF3 ELF3 O BCL6 ETV4 ETV4 BCL6 S SOX5 AD2A R CUX1 CUX1 THRB THRA THRB THRA RXRA RXRB RARA RERE RXRA RERE RXRB RARA MIXL1 MIXL1 G GATA4 ZG ZGPAT S SOX13 NR2F2 NR2F2 FOSL2 FOSL2 TEAD1 TEAD1 CEB CEBPA HNF4A MYBL2 MYBL2 HNF4A P PROX1 RREB1 RREB1 HNF4G HNF4G NC NCOA2 ARID5B ZNF219 ZNF219 ARID5B AT ZSCAN9 ZSCAN9 HMG20A HMG20A G GATAD2A ZKSCAN8 ZKSCAN8 0.5 0.5 NuRD Components

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0 Supplementary Figure 10. A. 37 non-FOX TFs with a called Forkhead moEf, with heatmap denoEng fracEon of called peaks with both a primary (matched to speciﬁc TF) moEf and a FOX moEf, with a primary moEf but not FOX moEf, with a FOX moEf but no primary moEf, and with neither a primary nor a FOX moEf. The eight TFs with gray boxes do not have a known primary moEf. B. Peak overlaps between the 37 TFs and six FOX factors for which we obtained ChIP-seq data; bar plots represent distribuEon of all FOX overlaps for each of the 37 factors. C. Same as B, but normalized for peak counts of each of the 37 factors. D. Same as C, but clustered verEcally, revealing NuRD component clustering. Boxes indicate quarEles, whiskers are drawn to minimum/maximum values. RERE_FLAG 1 HHEX_FLAG ZNF511_FLAG ZSCAN9_FLAG ZBTB21_FLAG ZNF3_FLAG ATF1_FLAG SRF SP2 ZNF274 SMC3 ZNF143 RFX5 EZH2 E2F7_iso1_FLAG NPAS2_iso2_FLAG ZNF544_iso1_FLAG 0.5 ZNF334_iso1_FLAG PBX2_FLAG PAF1_v1_FLAG ZC3H4_FLAG RAD21_FLAG CTCF DMAP1_FLAG GATAD1_FLAG KMT2B_FLAG MAZ SP1_FLAG MXI1 CHD2 SIX4_iso1_FLAG ZNF792_FLAG ZFP64_FLAG MIER2_FLAG 0 HMG20B_v2_FLAG HMGXB3_FLAG NR3C1 ZBTB33 CUX1 BRCA1 ZNF48_FLAG TGIF2_FLAG MXD4_FLAG HCFC1 NR2C2 HLF_FLAG NFYC_FLAG MAFK MAFF FOXP1_FLAG JUND -0.5 JUN GMEB2_FLAG ZNF7_iso2_FLAG RFX3_iso1_FLAG CEBPZ SIN3B SUZ12 THRB_FLAG NFE2L2 KAT2B BMI1 NCoA2_FLAG BACH1 IRF3 BRD4 RING1 ZNF384 TCF7L2 FOXA3_FLAG RXRB_FLAG SOX13_iso1_FLAG BCL6_iso1_FLAG TEAD1_FLAG TCF7 FOXO1 CEBPG_FLAG CEBPB CEBPA_FLAG NFIL3_FLAG ZNF644_FLAG NR2F1_FLAG FOS TFE3_FLAG MLX_FLAG CREB1_FLAG TBL1XR1 ARID3A PPARG_v1_FLAG MIXL1_FLAG KDM6A NFIA_v1_FLAG POLR2AphosphoS2 KLF16_FLAG DRAP1_FLAG TFAP4 CREM TBP NRF1 BHLHE40 USF2 USF1 ELF3_FLAG ASH2L HMG20A_FLAG KDM1A MYBL2 FOXA2 SOX5_FLAG FOXK1_FLAG ZNF281_FLAG PAXIP1_iso1_FLAG NR2F6_FLAG HNF4A ZNF219_FLAG GATAD2A_FLAG EP300 ARID5B_FLAG RARA_FLAG PROX1 HNF4G SOX13 HNF1A GATA4 FOXA1 NFIC HOMEZ_FLAG ZNF652_FLAG RCOR2_FLAG TEAD4 ZEB1 DNMT3B_FLAG ATF4_FLAG RFXANK_FLAG KLF9_FLAG ZHX3_iso1_FLAG KLF11_FLAG ZCCHC11_FLAG SSRP1_FLAG MBD1_v2_FLAG MBD1_v1_FLAG ZNF12_FLAG ZKSCAN8_FLAG JARID2_iso1_FLAG KDM3A_FLAG CIZ1_FLAG SLC30A9_FLAG ZFP1_v1_FLAG ZGPAT_FLAG ARID4B_FLAG POLR2AphosphoS5 POLR2A SIN3A THAP11_FLAG KAT8_FLAG GABPA_FLAG REST KLF10 MBD4 CEBPD HSF1 RUVBL1 ZNF189 RCOR1 NR1H2 ZNF335_FLAG UBP1_FLAG ZNF580_FLAG ZSCAN29_iso1_FLAG THRA_iso1_FLAG KLF6_v2_FLAG TEAD3_FLAG ZMYM3 RXRA SP5_FLAG ZBTB7A ATF3 ZBTB26_FLAG MTA1_FLAG TFDP1_FLAG SREBF1 KDM2A_FLAG MXD3_v1_FLAG GABPB1_v2_FLAG YY1 ERF_FLAG TAF1 SMAD4_iso1_FLAG SAP130_FLAG MAX FOSL2 ETV4 ZHX2 ESRRA ZNF331_FLAG PRDM10_FLAG HMGXB4_FLAG HBP1_FLAG KAT7_FLAG MYC RREB1_iso2_FLAG CBX5_FLAG CBX1 HDAC2 NR2F2 TCF25_FLAG MIER3_FLAG TCF12 RERE_FLAG HHEX_FLAG ZNF511_FLAG ZSCAN9_FLAG ZBTB21_FLAG ZNF3_FLAG ATF1_FLAG SRF SP2 ZNF274 SMC3 ZNF143 RFX5 EZH2 E2F7_iso1_FLAG NPAS2_iso2_FLAG ZNF544_iso1_FLAG ZNF334_iso1_FLAG PBX2_FLAG PAF1_v1_FLAG ZC3H4_FLAG RAD21_FLAG CTCF DMAP1_FLAG GATAD1_FLAG KMT2B_FLAG MAZ SP1_FLAG MXI1 CHD2 SIX4_iso1_FLAG ZNF792_FLAG ZFP64_FLAG MIER2_FLAG HMG20B_v2_FLAG HMGXB3_FLAG NR3C1 ZBTB33 CUX1 BRCA1 ZNF48_FLAG TGIF2_FLAG MXD4_FLAG HCFC1 NR2C2 HLF_FLAG NFYC_FLAG MAFK MAFF FOXP1_FLAG JUND JUN GMEB2_FLAG ZNF7_iso2_FLAG RFX3_iso1_FLAG CEBPZ SIN3B SUZ12 THRB_FLAG NFE2L2 KAT2B BMI1 NCoA2_FLAG BACH1 IRF3 BRD4 RING1 ZNF384 TCF7L2 FOXA3_FLAG RXRB_FLAG SOX13_iso1_FLAG BCL6_iso1_FLAG TEAD1_FLAG TCF7 FOXO1 CEBPG_FLAG CEBPB CEBPA_FLAG NFIL3_FLAG ZNF644_FLAG NR2F1_FLAG FOS TFE3_FLAG MLX_FLAG CREB1_FLAG TBL1XR1 ARID3A PPARG_v1_FLAG MIXL1_FLAG KDM6A NFIA_v1_FLAG POLR2AphosphoS2 KLF16_FLAG DRAP1_FLAG TFAP4 CREM TBP NRF1 BHLHE40 USF2 USF1 ELF3_FLAG ASH2L HMG20A_FLAG KDM1A MYBL2 FOXA2 SOX5_FLAG FOXK1_FLAG ZNF281_FLAG PAXIP1_iso1_FLAG NR2F6_FLAG HNF4A ZNF219_FLAG GATAD2A_FLAG EP300 ARID5B_FLAG RARA_FLAG PROX1 HNF4G SOX13 HNF1A GATA4 FOXA1 NFIC HOMEZ_FLAG ZNF652_FLAG RCOR2_FLAG TEAD4 ZEB1 DNMT3B_FLAG ATF4_FLAG RFXANK_FLAG KLF9_FLAG ZHX3_iso1_FLAG KLF11_FLAG ZCCHC11_FLAG SSRP1_FLAG MBD1_v2_FLAG MBD1_v1_FLAG ZNF12_FLAG ZKSCAN8_FLAG JARID2_iso1_FLAG KDM3A_FLAG CIZ1_FLAG SLC30A9_FLAG ZFP1_v1_FLAG ZGPAT_FLAG ARID4B_FLAG POLR2AphosphoS5 POLR2A SIN3A THAP11_FLAG KAT8_FLAG GABPA_FLAG REST KLF10 MBD4 CEBPD HSF1 RUVBL1 ZNF189 RCOR1 NR1H2 ZNF335_FLAG UBP1_FLAG ZNF580_FLAG ZSCAN29_iso1_FLAG THRA_iso1_FLAG KLF6_v2_FLAG TEAD3_FLAG ZMYM3 RXRA SP5_FLAG ZBTB7A ATF3 ZBTB26_FLAG MTA1_FLAG TFDP1_FLAG SREBF1 KDM2A_FLAG MXD3_v1_FLAG GABPB1_v2_FLAG YY1 ERF_FLAG TAF1 SMAD4_iso1_FLAG SAP130_FLAG MAX FOSL2 ETV4 ZHX2 ESRRA ZNF331_FLAG PRDM10_FLAG HMGXB4_FLAG HBP1_FLAG KAT7_FLAG MYC RREB1_iso2_FLAG CBX5_FLAG CBX1 HDAC2 NR2F2 TCF25_FLAG MIER3_FLAG TCF12

Supplementary Figure 11. Read count correlaEons between all 208 assayed factors, mean centered and squared, with unsupervised clustering. Motif Co−occurence Heatmap

Cis_Region Cis_Region

GATAD2A.2 ARID5B.1 ZNF219.1 Promoter_like PROX1.2 SOX5.2 ELF3.2 RARA.2 RXRB.2 GATA4.2 ETV4.2 Enhancer_like SP5.2 10 MLX.3 ERF.2 MIXL1.1 ZGPAT.1 SOX13.2 MYBL2.3 Promoter_and_Enhancer_like HMG20A.1 NCOA2.1 NFIC.1 FOXA3.1 FOXA3.2 FOXA1.1 FOXA2.1 FOXA2.2 Heterochromatin_and_Repressed CEBPA.2 TEAD1.3 HNF4A.3 HNF4G.3 NR2F2.2 RXRA.2 THRB.3 CTCF_Bound HSF1.2 MBD4.3 5 NCOA2.4 CUX1.4 MIXL1.3 GATA4.1 MLX.4 HMG20A.3 HOMEZ.3 SP5.3 SOX13.3 TCF7L2.1 NFIC.2 RERE.2 ZKSCAN8.1 ZKSCAN8.3 PROX1.3 ZNF644.3 RXRB.3 ZNF219.2 ARID5B.3 GATAD2A.1 HMG20A.4 ARID3A.5 FOXP1.4 0 FOSL2.4 PROX1.1 GATAD2A.3 HNF4A.1 HNF4G.1 ARID5B.2 HNF4A.4 SP5.4 RXRA.1 RXRB.1 ZNF3.2 MTA1.2 ZNF331.2 ZNF331.4 CEBPD.1 ZBTB26.1 NFIC.3 TEAD4.1 NR2F1.5 NR2F2.4 ESRRA.1 RCOR2.2 −5 ZMYM3.1 FOSL2.5 FOXK1.1 ZSCAN9.2 CUX1.3 FOXO1.2 ZNF3.1 NFIL3.1 FOXP1.2 HNF1A.1 SOX5.1 NR2F2.1 PROX1.5 ZNF219.4 ARID5B.4 TCF7.1 HMG20A.2 SOX13.1 CUX1.1 ZNF12.1 ZNF12.2 ZNF12.3 RARA.1 −10 NR2F1.1 NR2F6.1 ATF4.1 TEAD1.1 TEAD3.1 FOXA1.2 FOXP1.1 JUN.2 FOSL2.1 FOSL2.2 FOS.4 ZNF644.4 HSF1.1 HSF1.4 ZMYM3.3 ZNF189.3 ZMYM3.2 ZNF384.1 ZNF384.3 JUN.1 JUND.1 THRB.1 ZNF644.2 ELF3.1 TFAP4.1 HLF.1 CEBPA.1 CEBPG.1 CEBPA.3 CEBPB.1 SRF.1 ZNF652.1 MAFF.1 MAFK.1 BACH1.3 BACH1.1 NFE2L2.1 HSF1.3 NR1H2.1 ETV4.1 TGIF2.2 ZEB1.1 REST.1 ZNF331.1 ZNF48.2 DRAP1.3 PRDM10.1 TFE3.1 ZFP64.2 REST.2 REST.4 RAD21.2 CTCF.1 CTCF.2 HMGXB4.1 TCF25.1 RAD21.1 ZNF384.2 ZNF189.2 DNMT3B.1 ERF.1 ZNF792.1 TGIF2.1 ZNF511.1 ZSCAN9.1 KLF11.1 SLC30A9.1 KLF16.1 ZNF580.1 HBP1.1 UBP1.1 RCOR1.1 ZCCHC11.1 ZNF189.1 ZFP64.1 ZFP64.3 SP5.1 MLX.1 ZNF281.1 MYC.1 ZBTB7A.1 ZNF143.1 ZNF143.2 KLF10.1 KLF10.2 THAP11.1 THAP11.2 NR2C2.2 ZBTB21.1 MXD4.1 MXI1.1 BHLHE40.1 MAX.1 ZNF335.2 NRF1.1 NRF1.2 TFDP1.1 MAZ.1 MAZ.4 ATF3.3 GMEB2.2 NR2C2.4 GABPA.1 ARID4B.1 GABPA.2 GATAD1.1 MXI1.3 FOS.2 FOS.3 ATF3.1 ATF3.5 SREBF1.1 TFDP1.3 KLF9.1 DRAP1.1 SP1.1 SP2.2 TCF25.3 MXI1.2 YY1.1 ZHX2.2 CREM.1 ATF1.1 CREB1.1 GATAD1.2 ZHX2.1 NR2C2.1 RCOR1.2 ZBTB33.1 RFX5.1 RFXANK.1 PBX2.3 USF1.5 USF1.1 USF2.1 RFX5.2 RFXANK.2 SREBF1.5 NFYC.1 NFYC.2 IRF3.1 IRF3.2 PBX2.4 FOS.1 SP2.1 SP2.3 FOS.5 NFYC.5 PBX2.1 SP2.4 SP5.1 DRAP1.1 SP1.1 RCOR2.2 ZMYM3.2 TFAP4.1 ZNF331.1 ZSCAN9.2 GATA4.2 NCOA2.1 SOX13.2 THRB.3 FOXO1.2 NR2F2.2 RXRA.2 HNF4A.3 HNF4G.3 FOSL2.4 JUN.2 HLF.1 NFIL3.1 HMG20A.4 PROX1.3 NFIC.2 RERE.2 SP5.3 NR2F2.1 NR2F2.4 MLX.4 HMG20A.3 HOMEZ.3 NR2F1.5 NFIC.3 SP5.4 JUN.1 JUND.1 ZNF384.1 HNF1A.1 SOX13.3 TCF7L2.1 GATA4.1 SOX13.1 ARID3A.5 FOXP1.4 NCOA2.4 FOSL2.5 MAFF.1 MAFK.1 BACH1.3 BACH1.1 NFE2L2.1 THRB.1 FOS.4 ZNF644.4 TEAD4.1 CEBPD.1 ZNF644.3 ERF.1 KLF16.1 REST.2 TCF25.1 ZSCAN9.1 TGIF2.1 ZNF580.1 HBP1.1 KLF11.1 SLC30A9.1 ZNF792.1 UBP1.1 ZNF511.1 ZNF48.2 ZNF331.4 TGIF2.2 ZNF331.2 DRAP1.3 ZNF652.1 ZBTB26.1 ZNF644.2 ZBTB7A.1 MXD4.1 BHLHE40.1 MLX.1 ZBTB21.1 MXI1.2 GATAD1.2 ZHX2.1 RFX5.1 RFXANK.1 FOS.2 FOS.3 ATF3.1 ATF3.5 NR2C2.1 NR2C2.4 ATF3.3 GMEB2.2 MXI1.1 ZBTB33.1 ZMYM3.3 IRF3.1 IRF3.2 ZNF3.2 DNMT3B.1 REST.1 REST.4 RCOR1.1 ZCCHC11.1 PRDM10.1 SRF.1 ZEB1.1 ZNF384.2 ZFP64.1 ZFP64.3 CUX1.3 SREBF1.5 HSF1.3 NR1H2.1 HSF1.1 HSF1.2 HSF1.4 ZNF189.2 ZNF189.3 ZFP64.2 MYC.1 ZNF189.1 TFDP1.3 SREBF1.1 SP2.2 SP2.3 NR2C2.2 RCOR1.2 TCF25.3 ESRRA.1 ZMYM3.1 ZNF3.1 MTA1.2 ZKSCAN8.1 ZNF12.1 ATF4.1 ZKSCAN8.3 MBD4.3 CUX1.1 CUX1.4 ZNF12.2 ZNF12.3 TFE3.1 USF1.5 CREM.1 ATF1.1 CREB1.1 MXI1.3 YY1.1 ZHX2.2 KLF10.1 KLF10.2 ZNF143.1 ZNF143.2 MAX.1 TFDP1.1 KLF9.1 ZNF335.2 RFX5.2 RFXANK.2 PBX2.1 SP2.4 NFYC.5 PBX2.3 FOS.1 FOS.5 PBX2.4 SP2.1 NFYC.1 NFYC.2 MAZ.1 MAZ.4 GABPA.1 GABPA.2 ARID4B.1 GATAD1.1 USF1.1 USF2.1 NRF1.1 NRF1.2 HMGXB4.1 THAP11.1 THAP11.2 RAD21.1 CTCF.1 CTCF.2 ELF3.1 ETV4.1 ZNF281.1 RAD21.2 FOXA1.2 FOXA2.1 FOXA2.2 CEBPA.2 SOX5.2 TEAD1.3 FOXK1.1 ETV4.2 MLX.3 MYBL2.3 HMG20A.1 NFIC.1 ERF.2 SP5.2 FOXP1.1 MIXL1.1 PROX1.2 RXRB.3 ARID5B.3 GATAD2A.1 ZNF219.2 CEBPB.1 CEBPG.1 CEBPA.1 CEBPA.3 ZNF384.3 FOXP1.2 TCF7.1 HMG20A.2 SOX5.1 PROX1.5 ZNF219.4 NR2F1.1 RXRA.1 TEAD1.1 TEAD3.1 FOSL2.1 FOSL2.2 NR2F6.1 RARA.1 HNF4A.4 RXRB.1 ARID5B.4 MIXL1.3 GATAD2A.3 ARID5B.2 PROX1.1 HNF4A.1 HNF4G.1 ELF3.2 ZGPAT.1 RARA.2 RXRB.2 ARID5B.1 GATAD2A.2 ZNF219.1 FOXA1.1 FOXA3.1 FOXA3.2

Supplementary Figure 12. DirecEonal co-occurrence of moEfs in ChIP-seq called peaks. Metacluster 22 53 148 163 32 31 145 29 108 20 171 56 137 34 120 82 S2 SRF SP2 HLF TBP FOS JUN IRF3 MYC MAZ KLF9 MAX CTCF RFX5 RFX3 ATF4 ATF3 PAF1 USF1 USF2 ZNF3 ZNF7 PBX2 EZH2 ZHX3 ZHX2 NFYC MXI1 NRF1 BMI1 CUX1 HDF1 BRD4 HHEX CHD2 MAFF SMC3 SIN3B MAFK KLF11 KLF10 ZFP64 TCF12 EP300 CEBPZ RING1 SUZ12 ESRRA KAT2B HCFC1 BRCA1 NPAS2 NR2C2 NR3C1 GATA4 BACH1 RCOR1 JARID2 NR1H2 TCF7L3 GMEB2 NFE2L2 ZBTB21 ZBTB33 ZNFT43 ZNF274 ZNF334 ZNF281 ZNF544 ZNF384 ZBTB7A Phos RUVBL1 RFXANK TBL1XR1 HMG208 DNMT3B ZSCAN29 BHLHE40 ESRRA control POLR2A

Supplementary Figure 13. Example heatmaps showing factor enrichment in 16 key metaclusters. A

B C Metacluster 32 Metacluster 32

Supplementary Figure 14. A. SOMs for FOXA1, FOXA2, HNF4A, and EP300. B. Example decision tree showing presence/absence of factors for Metacluster 32. C. GREAT analysis of Metacluster 32 assigned genes likely regulated in this metacluster, and GO term analysis for these genes. chr2:228,314,800-228,317,799 1 kb H3K4me1 H3K27ac BACH1 SMC3 JUN CUX1 ARID3A CEBPB NFE2L2 JUND TBP RNA Pol2 MAFF MAFK p300 USF2 SP1 KLF10 FOXO1 ATF1 CREB1 ATF4 NR2F2 MAX ZNF219 MBD1_v2 DNMT3B TCF25 SSRP1 FOXA3 HLF GATAD2A SAP130 NR2F6 CEBPG

TEAD3

ARID5B FOXK1 NFIL3 SOX13 BCL6 ZNF331 DNase

Supplementary Figure 15. Example of genomic site with many associated factors. Each track shows aligned ChIP-seq reads, and is slightly oﬀset to be9er show peaks for each experiment. A. B.

Supplementary Figure 16. A. Enrichment of biological pathways at HOT regions near enhancers or promoters. B. Co-binding analysis of factors bound to HOT regions classiﬁed as enhancer-like by IDEAS. C. Co-binding analysis of factors bound to HOT regions classiﬁed as promoter-like by IDEAS. ConnecEng lines indicate high overlap of peaks, with percentages shown by color of lines. For enhancers: yellow = 75-79%, green = 80-89%, red = 90+%. For promoters: yellow = 70-79%, green = 80-89%, red = 90+%. Supplementary Figure 17. Increasing numbers of factors bound at genomic sites (less than 2 kb in size) associate with decreasing distance to nearest TSS (leT) and with increasing expression of nearest gene (right). Boxes represent middle two quarEles, whiskers are 1.5X inter-quarEle range. TSS = TranscripEon Start Site, FPKM = Fragments Per Kilobase of transcript per Million reads. GERP Highly Constrained Element overlap with Number of TFs bound

Supplementary Figure 18. Increasing numbers of factors bound at genomic sites correlate with increased evoluEonary constraint as measured by GERP (Genomic EvoluEonary Rate Profiling) showing incremental fracEon overlap of highly constrained elements with factor-associated sites, for both promoter regions (red) and enhancer regions (orange). Boxes indicate quarEles, whiskers are drawn to maximum value. Supplementary Figure 19. Number of unique DNase PIQ footprints (y- axis) plo9ed by sites with varying number of associated factors (x-axis), plo9ed for varying thresholds of PIQ purity scores (upper leT: > 0.7; upper right: > 0.8; lower leT: > 0.9; lower right: > 0.99). Supplementary Figure 20. DistribuEon of SVM classifier scores (y-axis) in sites with varying numbers of associated factors (x-axis). The scores remain relaEvely constant across sites and are significantly higher than the scores of classifier values in matched null sites. A

B Mean PR−AUC=0.66 0.8 0.6 0.4 0.2 CR/CF DBF

Supplementary Figure 21. SVM PR-AUC (Precision Recall Area Under Curve) scores for chromaEn regulators and cofactors (CR/CF) and for DNA-binding transcripEon factors (DBF) A) MoEf Level mean PR-AUC (0.74) B) Peak Level mean PR-AUC (0.66) Supplementary Figure 22. Number of sites (y-axis) with measured number of TFs (x-axis) with classifier values in the top 5% of all classifier values (steel blue) or with classifier values in the bo9om 75% of all classifier values (red) in highly bound regions, based on SVM scores of factor peaks associated to highly bound regions. A

B Supplementary Figure 23. Number of sites (y-axis) with measured number of TFs (x-axis) with classifier values in the top 5% of all classifier values (Steel blue) or with classifier values in the bo9om 75% of all classifier values (red). A. DistribuEon in HOT sites with >70 associated TFs. B. DistribuEon in sites with 2-10 associated TFs. C. DistribuEon in random set of enhancers with any number of associated TFs (0 or more). C A

Supplementary Figure 23. Degree of MoEf enrichment in highly bound regions for all HepG2 expressed factors with available moEfs (n=365) for 3 Categories A) Top 3 moEfs enriched in highly bound sites with 50+ TF (highest p-value = 3.9e-146) B) Top 3 moEfs in Enhancer with 2-10 TFs (highest p-value = 1.8e-17) C) Top moEf in Random Genome Enhancer with 0+ TFs (highest p-value=6.9e-3) Mean SVM Score Distribution 0.4

− HOTsites NonHOTsites Random Enh 0.6 −

0.8 ● − Mean SVM Score for sites(n=5676) Mean SVM Score for 1.0 −

●

● ● ● ● 1.2 − HOT Sites(>70) NonHOT Sites(2−10) Random Genome

Supplementary Figure 24. DistribuEon of all SVM scores (y-axis) for HOT sites with >70 associated TFs (red), for sites with 2-10 associated TFs (green), and for random enhancer sites with 0+ TFs (blue). Supplementary Figure 25. Pie chart showing fracEon of HOT sites in which each factor has the highest SVM classiﬁer value, indicaEng the strongest moEf present.