Supplementary Material

Supplementary material

Jan Grau, Marcel H Schulz, Florian Schmidt

1 A NFATC1 (GM12878) ZNF274 (K562) 2 2 1 1 0 0 CpG CpG MH MH 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20

0.00 0.04 0.08 −0.06 −0.02 0.02 0.00 0.02 0.04 −0.025 −0.010 0.005 CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH) p = 7.34e−09 B ZBTB33 74:86

● 0.3 ●

● ● ● 0.2 ●

● ● ● 0.1 ● ● Train GM12878 HepG2 K562 liver ●● ● ●● ● ● ●● ●● ●● ●● ● ● ● ●● ● ●● ● ●●● ● 0.0 ● ● ● GM12878 ● HepG2 ● K562 ● liver −0.1 Test

−0.2 PWM.methyl − PWM.hg38 PWM.methyl

C GM12878 D liver 2 2 1 1 0 0 CpG CpG MH MH 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20

0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 −1.5 −1.0 −0.5 0.0 CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH)

Figure 1: Methylation sensitivity is positive for TFs with known preference for methylated DNA. (A) For NFATC1, we find methylation sensitivity values that are clearly positive at some positions (8,12), but also clearly negative at other positions, while the amplitude of methylation sensitivity and CpG content is rather low. Similarly, we find methylation sensitivity values that are clearly positive at some positions (especially position 10), but also clearly negative at other positions for ZNF274. (B) For ZBTB33, we include within cell type and across cell type comparisons in a single plot, which compares the performance of a methylation-aware PWM model with the corresponding PWM model learned on the original hg38 genome. For within cell type performance, we find an improvement when including methylation information for GM12878 (dark violet dots), HepG2 (blue triangles), and liver (yellow crosses), whereas results for K562 (green squares) are rather mixed. Across cell types , the motif trained on methylation-aware GM12878 data works particularly well for K562 (green dots), but less well for HepG2 (blue dots) and liver (yellow dots) data. In turn, the motif trained on methylation-aware liver data works well for HepG2 (blue crosses) but not for GM12787 (dark violet crosses) and K562 (green crosses) data. (C) Non-canonical motif discovered for ZBTB33 in GM12878 cells with clear methylation preference, which also performed well on K562 data. For K562, a strong methylation preference for ZBTB33 has been reported before (1). (D) The motif discovered from liver data resem- bles the canonical ZBTB33 motif, but shows a negative effect of binding site methylation. This motif also worked well for HepG2 data. FOXA1 FOXA2 FOXK2 2 2 2 1 1 1 0 0 0 CpG CpG CpG MH MH MH 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20

0.000 0.005 0.010 0.015 −0.025 −0.015 −0.005 0.000 0.005 0.010 0.015 −0.015 −0.005 0.00 0.04 0.08 −0.15 −0.05 HepG2 CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH) 2 2 1 1 0 0 CpG CpG MH MH 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20

0.000 0.010 0.020 −0.025 −0.010 0.005 0.00 0.10 0.20 −0.4 −0.2 0.0 K562 CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH) 2 2 1 1 0 0 CpG CpG MH MH 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20

0.000 0.006 0.012 −0.012 −0.006 0.000 0.000 0.005 0.010 0.015 −0.015 −0.005 liver CpG Methylation sensitivity (MH) CpG Methylation sensitivity (MH)

Figure 2: Methylation sensitivity may diﬀer between members of a TF family. While methylation sensitivity of FOXA1 and FOXA2 is highly similar in all cell types, that of FOXK2 is noticeably diﬀerent, although the motifs of all three TFs appear to be highly similar. K562 HepG2 GM12878 0.02 0.5 0.02 0.5 0.02 0.5

1 3 5 7 9 11 13 15 17 19 5 10 15 20 1 3 5 7 9 11 13 15 17 19 5 10 15 20 1 3 5 7 9 11 13 15 17 19 5 10 15 20 1.0 0.8 0.8 0.8 0.6 0.6 0.6 CpG CpG CpG 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 0.0 0.0 0.0 N = 9064 N = 60311 N = 56174 −0.2 −0.5 −0.5 −0.4 −1.0 −1.0 −0.6 −1.5 −1.5 −0.8 −2.0 Methylation sensitivity (MH) Methylation sensitivity (MH) Methylation sensitivity (MH) Methylation −2.0 −1.0 −2.5

5 10 15 20 5 10 15 20 5 10 15 20

1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 17 19

Figure 3: Dependency logos and profiles of methylation sensitivity for JUND in three cell types (K562, HepG2, GM12878). In all three cases, we consistently find both, the long-spacer and the short-spacer variant, of the binding motif, with a specific pattern of methylation sensitivity, which is prominent only for the long-spacer variant. Table 1: List of the Encode IDs of all ChIP-seq data sets used in this study for the four cell types GM12878, HepG2, K562, and liver. TF GM12878 K562 HepG2 liver AFF1 ENCFF489SKQ, ENCFF869BYK AGO1 ENCFF100VYA ENCFF627BHP AGO2 ENCFF465FII ARHGAP35 ENCFF089PKE ARID1B ENCFF249TYS ARID2 ENCFF344MKI ARID3A ENCFF003VDB ENCFF757OML ENCFF247GXE ARNT ENCFF758RQJ ENCFF447FIO, ENCFF616WXJ ENCFF655EFA, ENCFF913AQF ASH1L ENCFF958YSG ASH2L ENCFF096XRG ENCFF638IUM ATF2 ENCFF210HTZ ENCFF803FHN ENCFF089BQU ATF3 ENCFF467WOR, ENCFF137OEY ENCFF146URA, ENCFF937OKC ENCFF782SGI ATF4 ENCFF182MNO ATF7 ENCFF495PWL ENCFF371SJR ENCFF498YGH ATM ENCFF906FVB BACH1 ENCFF725YZH BATF ENCFF832YIE BCL11A ENCFF383HAY BCL3 ENCFF247MHT BCLAF1 ENCFF587BJK ENCFF054DTJ, ENCFF506PXL ENCFF094XPM, ENCFF496YJC BCOR ENCFF186JKG BHLHE40 ENCFF370ZNL ENCFF477JTV ENCFF361YXC, ENCFF863ATX BMI1 ENCFF592LPO ENCFF352DRR BRCA1 ENCFF005JKU ENCFF652NES ENCFF897ETK BRD4 ENCFF806CQB ENCFF736GHL BRD9 ENCFF411RMT CBFA2T2 ENCFF419PEK CBFA2T3 ENCFF153IFH CBFB ENCFF070SOX CBX1 ENCFF163FLA CBX2 ENCFF501QII CBX3 ENCFF552QOA ENCFF951BQB CBX5 ENCFF417SVR ENCFF403TAE CC2D1A ENCFF180TUM Table 1: (continued) TF GM12878 K562 HepG2 liver CCAR2 ENCFF704PGT ENCFF039LHY CDC5L ENCFF384ALH CEBPB ENCFF786YYI ENCFF321KQD ENCFF862DXR, ENCFF915ZYE CEBPBZ ENCFF797OWK CEBPZ ENCFF243GOG ENCFF195BKI CHAMP1 ENCFF646MEF, ENCFF919KNQ CHD1 ENCFF863CTN CHD2 ENCFF546AYN ENCFF181XMM CHD4 ENCFF249SIN ENCFF148ABR COPS2 ENCFF552EBC CREB1 ENCFF550TXR CREB3L1 ENCFF566HGU CREM ENCFF091YID ENCFF021XJN ENCFF290UGF CTBP1 ENCFF349UTF CTCF ENCFF356LIU ENCFF119XFJ, ENCFF543WTP ENCFF396BZQ, ENCFF519CXF, ENCFF843VHC CUX1 ENCFF567NFS ENCFF556HMX DACH1 ENCFF870LJV DDX20 ENCFF536LKB DEAF1 ENCFF532HCE DNMT1 ENCFF549TVW DPF2 ENCFF771IAW ENCFF217ZTP, ENCFF537VKZ E2F1 ENCFF134JLR, ENCFF445VTT E2F4 ENCFF687SFB E2F6 ENCFF533GSH E2F7 ENCFF013EHI E2F8 ENCFF412GFI ENCFF171WWF E4F1 ENCFF035GFS ENCFF752KNU EBF1 ENCFF249SVT EED ENCFF023ALY EGR1 ENCFF175VSS, ENCFF617JQS, ENCFF375RDB, ENCFF808WST ENCFF561OGS EHMT2 ENCFF682XPD ENCFF413RQL EKL1 ENCFF432AQP ELF1 ENCFF948CPI ENCFF617ZLL ENCFF840RWO Table 1: (continued) TF GM12878 K562 HepG2 liver ELF4 ENCFF539SXG ELK1 ENCFF119SCQ EP300 ENCFF510FUM ENCFF755HCK ENCFF674QCU, ENCFF806JJS EP400 ENCFF225BXA ESRRA ENCFF722LJP ENCFF592GWM ETS1 ENCFF980VOD ENCFF461PRP ENCFF128TUP ETV4 ENCFF710CRT ETV6 ENCFF116AMK ENCFF426GSY, ENCFF658SGJ EWSR1 ENCFF560CYG EZH2 ENCFF615NYO ENCFF504QZJ FIP1L1 ENCFF084DTV ENCFF031LBW FOSL1 ENCFF087MFG FOSL2 ENCFF054ESU FOXA1 ENCFF765NAN ENCFF152BOT, ENCFF324QGE, ENCFF367TQC, ENCFF951VPZ ENCFF872MGU FOXA2 ENCFF184NAC, ENCFF168JLI, ENCFF259BJR ENCFF293LRQ FOXK2 ENCFF990MTR ENCFF490EQR ENCFF315CHX FOXM1 ENCFF778PWE FOXP1 ENCFF029UJC FUS ENCFF688ARM ENCFF216YZI GABPA ENCFF946ACA ENCFF124HAC ENCFF054HJA ENCFF280YAF, ENCFF344XWK GABPB1 ENCFF700DXR GATA1 ENCFF148JKK GATA2 ENCFF173TXA GATA4 ENCFF097OXR GATAD2A ENCFF950ZWP GATAD2B ENCFF298AIX ENCFF569CMJ GMEB1 ENCFF678VPQ GTF2F1 ENCFF478HYJ, ENCFF599TWF, ENCFF713ALB, ENCFF876GXQ ENCFF843UHP HCFC1 ENCFF722QBB ENCFF167RXK ENCFF485SRU HDAC1 ENCFF188TBM, ENCFF069KPS ENCFF557WXK, ENCFF661VOO, ENCFF758PGF Table 1: (continued) TF GM12878 K562 HepG2 liver HDAC2 ENCFF299UPZ ENCFF363GSV, ENCFF182XZZ, ENCFF519RWJ, ENCFF589GSN ENCFF618YRQ HDAC3 ENCFF742LSD HDAC6 ENCFF295GBP ENCFF109EXK HDGF ENCFF442WRJ ENCFF297WMY, ENCFF575WFB HES1 ENCFF010OOE HMBOX1 ENCFF718DFX HNF1A ENCFF800QTO HNF4A ENCFF072CXB ENCFF837QHJ, ENCFF905JAC HNF4G ENCFF086CTA ENCFF497MUF HNRNPH1 ENCFF844QFF ENCFF046NUR HNRNPK ENCFF984QUV ENCFF828KXG HNRNPL ENCFF984ESZ ENCFF039CUI HNRNPLL ENCFF662WPN ENCFF890KTX HNRNPUL1 ENCFF991ZSC ENCFF509YFF HSF1 ENCFF603BID IKZF1 ENCFF968NOG ENCFF785BTP, ENCFF969BZA ENCFF994OQH IKZF2 ENCFF088OLI ILF3 ENCFF368AAQ IRF2 ENCFF886EVL IRF3 ENCFF604AZX IRF4 ENCFF720YMW IRF5 ENCFF843HDK JUN ENCFF394CEC JUNB ENCFF478XNA ENCFF739XTO JUND ENCFF873DJD ENCFF213EYD ENCFF430PEI, ENCFF229COM ENCFF539GRW KAT2A ENCFF710ROZ KAT2B ENCFF091BEK KAT8 ENCFF207ZEK KDM1A ENCFF799KZP ENCFF483BRD, ENCFF768FGG ENCFF728KKP, ENCFF796VMI KDM4B ENCFF470RHZ, ENCFF955AOD KDM5A ENCFF334HKG KDM5B ENCFF668XLN KLF16 ENCFF379LKE Table 1: (continued) TF GM12878 K562 HepG2 liver KLF5 ENCFF417WPC L3MBTL2 ENCFF423LPW LCORL ENCFF611PIO LEF1 ENCFF134HQP, ENCFF697VRJ MAFF ENCFF498MGH ENCFF493TIR MAFK ENCFF186AWV ENCFF893SCL ENCFF171OJF, ENCFF770TZL MAX ENCFF270NAL ENCFF618VMC, ENCFF140PUO ENCFF669BQN ENCFF900NVQ MAZ ENCFF348STZ ENCFF144TBQ MBD2 ENCFF617QSK MCM2 ENCFF043HHG, ENCFF571REC MCM3 ENCFF672PYP MCM5 ENCFF603SXI, ENCFF658SJY MCM7 ENCFF159MQI, ENCFF288ZRD, ENCFF914ELA MEF2A ENCFF958GXF ENCFF310SMW MEF2B ENCFF623FAW MEF2C ENCFF830BRO MEIS2 ENCFF937UEE MGA ENCFF525MPI MIER1 ENCFF163YZB MITF ENCFF071NYD, ENCFF262TMM MLLT1 ENCFF125MEN ENCFF010AIG, ENCFF388LUX MNT ENCFF454QQD, ENCFF482JSR, ENCFF459DYU, ENCFF562FMQ ENCFF926CRV MTA1 ENCFF801KEW MTA2 ENCFF587POH ENCFF558XIL, ENCFF713ZVD MTA3 ENCFF661FMB ENCFF459XLR MXI1 ENCFF199HGX ENCFF243QTL MYB ENCFF402TSJ MYBL2 ENCFF905KOD MYC ENCFF492XUU MYNN ENCFF272LLG NBN ENCFF811VEN ENCFF516UWH Table 1: (continued) TF GM12878 K562 HepG2 liver NCOA1 ENCFF382RFJ, ENCFF474QDS, ENCFF589OOF NCOA2 ENCFF071SOH, ENCFF584SNZ NCOA4 ENCFF749HKV NCOA6 ENCFF438BWN NCOR1 ENCFF856HUK ENCFF616RSZ NEUROD1 ENCFF755APC NFATC1 ENCFF138ZBJ NFATC3 ENCFF704PDA ENCFF082EPO, ENCFF430JFH NFE2 ENCFF312XHI NFE2L2 ENCFF882YLO NFIC ENCFF480WDX ENCFF092TVM NFKB ENCFF162TPR NFRKB ENCFF158FUG, ENCFF779KIS NFXL1 ENCFF860IXB ENCFF329STX NFYA ENCFF278GJK NFYB ENCFF510NDO NONO ENCFF515YFU, ENCFF108IZQ, ENCFF823CQK ENCFF420QKI NR0B1 ENCFF305OOU NR2C1 ENCFF462AKP ENCFF023XHV NR2C2 ENCFF434HVY ENCFF791ZPU NR2F1 ENCFF531KOV ENCFF363IQN NR2F2 ENCFF118HUH ENCFF379TVQ NR2F6 ENCFF194VBK ENCFF350CKI NR3C1 ENCFF315MUH, ENCFF821YMC NRF1 ENCFF652BRY ENCFF543STN, ENCFF313RFR, ENCFF626VDA, ENCFF418DKQ ENCFF782YFS NUFIP1 ENCFF885JMZ PAX5 ENCFF196JGP PAX8 ENCFF992JWY PBX3 ENCFF926LHG PCBP1 ENCFF467RYH ENCFF487WAN PCBP2 ENCFF941XZW ENCFF642XRH PHB2 ENCFF988OXX ENCFF882RPA PHF20 ENCFF259HUS Table 1: (continued) TF GM12878 K562 HepG2 liver PHF21A ENCFF657UVA PHF8 ENCFF952YDR ENCFF202WIO PKNOX1 ENCFF335ADU ENCFF062VBB PLRG1 ENCFF873OHG PML ENCFF800QDU POU5F1 ENCFF814QPF PRDM10 ENCFF600HPZ PRPF4 ENCFF417RQZ ENCFF908QCS PTBP1 ENCFF917HXV ENCFF875ZPV PYGO2 ENCFF442XXV RAD21 ENCFF654EGO ENCFF093XOJ, ENCFF229WFR ENCFF874VFZ RAD51 ENCFF996NBR ENCFF740OPF ENCFF859MBC RB1 ENCFF034OSV ENCFF328QZM RBBP5 ENCFF687SSY ENCFF666PCE RBFOX2 ENCFF232ASB ENCFF871YRG RBM14 ENCFF465UMU RBM15 ENCFF563WDZ RBM17 ENCFF056OIG RBM22 ENCFF420IBN ENCFF305WYD RBM25 ENCFF102XVH RBM34 ENCFF670ILH RBM39 ENCFF503DIK ENCFF420ALF RCOR1 ENCFF470ZMK ENCFF968SUH ENCFF987VKU RELB ENCFF105YDI REST ENCFF313CII ENCFF023ZUW, ENCFF669XCW, ENCFF178WRO, ENCFF290ESJ ENCFF986RRJ ENCFF288XHG RFX1 ENCFF193PVX, ENCFF788CJF ENCFF905GXS RFX5 ENCFF259LNG ENCFF201YKU ENCFF059GWW RLF ENCFF599CBB RNF2 ENCFF349MSP, ENCFF380SYL ENCFF462AZY, ENCFF741CLJ, ENCFF820LKT RUNX1 ENCFF091MQJ, ENCFF545WXN RUNX3 ENCFF677QUK RXRA ENCFF313BDA ENCFF105TFM ENCFF201KGJ SAFB ENCFF411YVY SAFB2 ENCFF087DKT SAP30 ENCFF103RHL Table 1: (continued) TF GM12878 K562 HepG2 liver SETDB1 ENCFF690WNQ SIN3A ENCFF050CYK ENCFF407VGB, ENCFF635YMI ENCFF802JAN SIN3B ENCFF543INR ENCFF193DQZ SIX5 ENCFF864TFH ENCFF247LOF SKI ENCFF035ZFO SKIL ENCFF903KEI ENCFF254QDM SMAD1 ENCFF987PGY ENCFF084BUP SMAD2 ENCFF186MFI SMAD5 ENCFF855SJG ENCFF069AAY SMARCA4 ENCFF361RWX, ENCFF703NAE, ENCFF868UOJ SMARCA5 ENCFF052STI ENCFF481TNF SMARCB1 ENCFF308QHX SMARCC2 ENCFF751ZVX ENCFF150NHK SMARCE1 ENCFF435SZS ENCFF210HAA SMC3 ENCFF572RPI ENCFF175UEE ENCFF035YWE SNIP1 ENCFF529BDW SNRNP70 ENCFF206MJS ENCFF858FBZ SOX13 ENCFF257QND SOX6 ENCFF431STY ENCFF944LNI SP1 ENCFF452LDK ENCFF175VXL, ENCFF433EFF, ENCFF735WMX ENCFF978TMH SPI1 ENCFF071ZMW ENCFF414ECK SREBF1 ENCFF777MYW SRF ENCFF182IFE SRSF3 ENCFF926XGK SRSF4 ENCFF122FVR SRSF7 ENCFF550VUN SRSF9 ENCFF217HAW ENCFF121PED STAT1 ENCFF323QQU STAT3 ENCFF923CHO STAT5A ENCFF517IXK SUPT20H ENCFF069YVD SUPT5H ENCFF721DPQ SUZ12 ENCFF440XZI, ENCFF239LRW ENCFF856HYC SYNCRIP ENCFF157IIV TAF1 ENCFF540AAP ENCFF234TBW ENCFF214OJW TAF15 ENCFF710LLF ENCFF718RXL TAF7 ENCFF852NOL Table 1: (continued) TF GM12878 K562 HepG2 liver TAF9B ENCFF223HDM TAL1 ENCFF078OUD, ENCFF475LFH TARDBP ENCFF668JHK ENCFF448YOS, ENCFF696QPP ENCFF641AXD, ENCFF909RMQ TBL1XR1 ENCFF392JWA ENCFF239WFN, ENCFF126KGW ENCFF868SWL TBP ENCFF896UZB ENCFF370YGS ENCFF534GKQ TBX21 ENCFF971VHK TBX3 ENCFF654KVO, ENCFF887DUY TCF12 ENCFF768VSH ENCFF912LXU, ENCFF299JYV, ENCFF952JIK ENCFF820PHL TCF7 ENCFF152RNE ENCFF512IAI ENCFF928MIN TCF7L2 ENCFF556FYF TEAD4 ENCFF547MLB TFAP4 ENCFF912SQI THAP1 ENCFF130TPD THRA ENCFF309DMZ THRAP3 ENCFF354UUL TRIM22 ENCFF830TFU ENCFF063GDN TRIM24 ENCFF063NXI, ENCFF950TOJ TRIM28 ENCFF168KHS, ENCFF623ELO, ENCFF996AMX TRIP13 ENCFF534VQL U2AF1 ENCFF482DRO ENCFF034KUO U2AF2 ENCFF134HBP ENCFF562ADR UBTF ENCFF295ZLM ENCFF345RRM, ENCFF403TAF USF1 ENCFF914IFQ USF2 ENCFF514SWA ENCFF425FVY WRNIP1 ENCFF514DDI XRCC3 ENCFF115PGE XRCC5 ENCFF929TWP YBX1 ENCFF500RBO ENCFF520DIY ENCFF332FUE YBX3 ENCFF508WCC YY1 ENCFF223MUF ENCFF024TJO, ENCFF177YDT ENCFF459TWF ENCFF635XCI ZBED1 ENCFF630FLK ENCFF388TYU ZBTB11 ENCFF913HCQ Table 1: (continued) TF GM12878 K562 HepG2 liver ZBTB2 ENCFF189WAO ZBTB33 ENCFF475DID ENCFF556STK ENCFF943WRA ENCFF727ZIT ZBTB40 ENCFF084IUW ENCFF088LZZ ZBTB5 ENCFF014KUI, ENCFF813GMP ZBTB7A ENCFF245LRG ENCFF953JQD ZBTB8A ENCFF328SSL ZC3H11A ENCFF478PGJ ZEB2 ENCFF553KIK, ENCFF808NWU ZFP36 ENCFF224WII ZFP91 ENCFF150ZBH ZHX1 ENCFF495BPY ZHX2 ENCFF964KDQ ZKSCAN1 ENCFF704VDI ENCFF721NEC ZMIZ1 ENCFF526PMI ZMYM3 ENCFF195IFB ENCFF769SEZ ZNF143 ENCFF153TQR ENCFF700GZI ZNF184 ENCFF855CUN ZNF184A ENCFF760EPB ZNF207 ENCFF676BIG ENCFF657ZXY ZNF217 ENCFF200SLC ZNF24 ENCFF313HBL ENCFF007EEV, ENCFF858WPR, ENCFF260CBQ, ENCFF904QAD ENCFF723JDW ZNF274 ENCFF323AWS, ENCFF498VQZ ZNF280A ENCFF074WRG ZNF282 ENCFF596JDS ENCFF482XNG ZNF316 ENCFF056SEM, ENCFF806GUF ZNF318 ENCFF082RIZ, ENCFF577LQR ZNF384 ENCFF942MDT ENCFF106YXG ENCFF950VAR ZNF407 ENCFF538GSS, ENCFF644XES ZNF592 ENCFF615DTQ ENCFF972UGK ZNF622 ENCFF777DVJ ZNF639 ENCFF008JJE, ENCFF404EVY ZNF687 ENCFF137BRA Table 1: (continued) TF GM12878 K562 HepG2 liver ZNF830 ENCFF951OSW, ENCFF979NKM ZPF36 ENCFF166GKK ZSCAN29 ENCFF214NJL ENCFF908ZLN, ENCFF979GFF ZTBTB40 ENCFF624WDI ZZZ3 ENCFF260NAX ENCFF945HJR

Table 2: Summary of the TFs under study. The first column specifies the TF and the following four columns indicate the availability of ChIP-seq data for the different cell types. In column “Methylation”, “NA” indicates that data sets have been available only for a single cell type, preventing across cell type predictions; “n” indicates that no significant increase in performance could be observed when including methylation information; “i” indicates inconsistent results between the different strategies for sampling negative training examples; “y” indicates consistent and significant increase in performance when including methylation information. The entries in column “Methyl. & Deps. have the same meaning but for increasing performance when considering methylation information and intra-motif dependencies. In column “Literature”, “-”, “+” and “s” indicate negative or positive influence of methylation or general methylation sensitivity, respectively; “new” indicates cases that have been identified as methylation sensitive in this study. TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature AFF1 x NA NA AGO1 x x n n AGO2 x NA NA ARHGAP35 x NA NA ARID1B x NA NA ARID2 x NA NA ARID3A x x x y n new ARNT x x x y n - (2) ASH1L x NA NA ASH2L x x n n ATF2 x x x i n - (3) ATF3 x x x y y - (3) ATF4 x NA NA - (4; 5) ATF7 x x x y n - (3) ATM x NA NA BACH1 x NA NA BATF x NA NA different motif (6) BCL11A x NA NA BCL3 x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature BCLAF1 x x x n n BCOR x NA NA BHLHE40 x x x y n - (3) BMI1 x x i n BRCA1 x x x y n new BRD4 x x i n BRD9 x NA NA CBFA2T2 x NA NA CBFA2T3 x NA NA CBFB x NA NA CBX1 x NA NA CBX2 x NA NA CBX3 x x n n CBX5 x x n n CC2D1A x NA NA CCAR2 x x i n CDC5L x NA NA CEBPB x x x n n +/- (4; 7; 8; 5) CEBPBZ x NA NA CEBPZ x x i n CHAMP1 x NA NA CHD1 x NA NA CHD2 x x i n CHD4 x x y n new COPS2 x NA NA CREB1 x NA NA - (3) CREB3L1 x NA NA CREM x x x y n - (3) CTBP1 x NA NA CTCF x x x n n - (2; 9; 10; 11) CUX1 x x i n - (3) DACH1 x NA NA DDX20 x NA NA DEAF1 x NA NA DNMT1 x NA NA DPF2 x x n n E2F1 x NA NA - (3) E2F4 x NA NA - (3) E2F6 x NA NA E2F7 x NA NA - (3) Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature E2F8 x x i n E4F1 x x i n EBF1 x NA NA EED x NA NA EGR1 x x n n EHMT2 x x i n EKL1 x NA NA ELF1 x x x y n - (4; 3) ELF4 x NA NA - (3) ELK1 x NA NA - (4; 3) EP300 x x x i n EP400 x NA NA ESRRA x x i n ETS1 x x x i y - (4) ETV4 x NA NA - (3) ETV6 x x i n EWSR1 x NA NA EZH2 x x n n FIP1L1 x x n n FOSL1 x NA NA FOSL2 x NA NA FOXA1 x x x y n new FOXA2 x x y y new FOXK2 x x x y n new FOXM1 x NA NA FOXP1 x NA NA FUS x x n n GABPA x x x x y n - (3) GABPB1 x NA NA GATA1 x NA NA + (4) GATA2 x NA NA +/- (4; 8) GATA4 x NA NA + (4) GATAD2A x NA NA GATAD2B x x i n GMEB1 x NA NA - (3) GTF2F1 x x i n HCFC1 x x x y y new HDAC1 x x n n HDAC2 x x x y n new HDAC3 x NA NA HDAC6 x x n n HDGF x x n n Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature HES1 x NA NA - (3) HMBOX1 x NA NA HNF1A x NA NA HNF4A x x y y new HNF4G x x y y new HNRNPH1 x x i n HNRNPK x x n n HNRNPL x x n n HNRNPLL x x i n HNRNPUL1 x x i n HSF1 x NA NA IKZF1 x x x i n IKZF2 x NA NA ILF3 x NA NA IRF2 x NA NA + (3) IRF3 x NA NA IRF4 x NA NA IRF5 x NA NA JUN x NA NA - (3) JUNB x x i n - (3) JUND x x x x y y - (3) KAT2A x NA NA KAT2B x NA NA KAT8 x NA NA KDM1A x x x i n KDM4B x NA NA KDM5A x NA NA KDM5B x NA NA KLF16 x NA NA + (3) KLF5 x NA NA L3MBTL2 x NA NA LCORL x NA NA LEF1 x NA NA MAFF x x n n MAFK x x x n n MAX x x x x y y s/- (2; 3) MAZ x x i n MBD2 x NA NA MCM2 x NA NA MCM3 x NA NA MCM5 x NA NA MCM7 x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature MEF2A x x n n MEF2B x NA NA MEF2C x NA NA MEIS2 x NA NA MGA x NA NA MIER1 x NA NA MITF x NA NA MLLT1 x x n n MNT x x y n s/- (2) MTA1 x NA NA MTA2 x x y n new MTA3 x x i n MXI1 x x i n MYB x NA NA MYBL2 x NA NA - (3) MYC x NA NA MYNN x NA NA NBN x x n n NCOA1 x NA NA NCOA2 x NA NA NCOA4 x NA NA NCOA6 x NA NA NCOR1 x x n n NEUROD1 x NA NA NFATC1 x NA NA + (3) NFATC3 x x y n + (3) NFE2 x NA NA NFE2L2 x NA NA NFIC x x i n NFKB x NA NA NFRKB x NA NA NFXL1 x x n n NFYA x NA NA s/- (2) NFYB x NA NA NONO x x y y new NR0B1 x NA NA NR2C1 x x n n NR2C2 x x y n new NR2F1 x x n n NR2F2 x x i n NR2F6 x x n n NR3C1 x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature NRF1 x x x y n - (2) NUFIP1 x NA NA PAX5 x NA NA PAX8 x NA NA + (3) PBX3 x NA NA PCBP1 x x n n PCBP2 x x n n PHB2 x x n n PHF20 x NA NA PHF21A x NA NA PHF8 x x n n PKNOX1 x x y n new PLRG1 x NA NA PML x NA NA POU5F1 x NA NA PRDM10 x NA NA PRPF4 x x n n PTBP1 x x n n PYGO2 x NA NA RAD21 x x x y y new RAD51 x x x y y new RB1 x x i n RBBP5 x x n n RBFOX2 x x n n RBM14 x NA NA RBM15 x NA NA RBM17 x NA NA RBM22 x x i n RBM25 x NA NA RBM34 x NA NA RBM39 x x n n RCOR1 x x x i n RELB x NA NA REST x x x x i n - (10) RFX1 x x n n RFX5 x x x i n RLF x NA NA RNF2 x x y n new RUNX1 x NA NA RUNX3 x NA NA - (8; 3) RXRA x x x i n SAFB x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature SAFB2 x NA NA SAP30 x NA NA SETDB1 x NA NA SIN3A x x x i n SIN3B x x i n SIX5 x x y n new SKI x NA NA SKIL x x i n SMAD1 x x i n SMAD2 x NA NA SMAD5 x x n n + (3) SMARCA4 x NA NA SMARCA5 x x i n SMARCB1 x NA NA SMARCC2 x x n n SMARCE1 x x n n SMC3 x x x i n SNIP1 x NA NA SNRNP70 x x n n SOX13 x NA NA SOX6 x x n n SP1 x x x y y +/- (3; 12; 13; 14) SPI1 x x n n SREBF1 x NA NA - (3) SRF x NA NA SRSF3 x NA NA SRSF4 x NA NA SRSF7 x NA NA SRSF9 x x n n STAT1 x NA NA + (4) STAT3 x NA NA STAT5A x NA NA + (4) SUPT20H x NA NA SUPT5H x NA NA SUZ12 x x i n SYNCRIP x NA NA TAF1 x x x y n new TAF15 x x n n TAF7 x NA NA TAF9B x NA NA TAL1 x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature TARDBP x x x i n TBL1XR1 x x x y n new TBP x x x n n TBX21 x NA NA TBX3 x NA NA TCF12 x x x n n TCF7 x x x i n TCF7L2 x NA NA TEAD4 x NA NA TFAP4 x NA NA THAP1 x NA NA THRA x NA NA THRAP3 x NA NA TRIM22 x x n n TRIM24 x NA NA TRIM28 x NA NA TRIP13 x NA NA U2AF1 x x n n U2AF2 x x n n UBTF x x n n USF1 x NA NA - (2; 3) USF2 x x y y - (3) WRNIP1 x NA NA XRCC3 x NA NA XRCC5 x NA NA YBX1 x x x n n YBX3 x NA NA YY1 x x x x y n different motif (15) ZBED1 x x n n - (3) ZBTB11 x NA NA ZBTB2 x NA NA - (3) ZBTB33 x x x x n n + (3; 16; 17; 18) ZBTB40 x x y n new ZBTB5 x NA NA ZBTB7A x x n n - (3) ZBTB8A x NA NA ZC3H11A x NA NA ZEB2 x NA NA ZFP36 x x n n ZFP91 x NA NA Table 2: (continued) TF GM12878 HepG2 K562 liver Methylation Methyl. & Deps. Literature ZHX1 x NA NA ZHX2 x NA NA ZKSCAN1 x x i n ZMIZ1 x NA NA ZMYM3 x x n n ZNF143 x x n n ZNF184 x NA NA ZNF184A x NA NA ZNF207 x x n n ZNF217 x NA NA ZNF24 x x x i n ZNF274 x NA NA + (3) ZNF280A x NA NA ZNF282 x x n n ZNF316 x NA NA ZNF318 x NA NA ZNF384 x x x i n ZNF407 x NA NA ZNF592 x x n n ZNF622 x NA NA ZNF639 x NA NA ZNF687 x NA NA ZNF830 x NA NA ZPF36 x NA NA ZSCAN29 x x i n ZTBTB40 x NA NA ZZZ3 x x i n Table 3: Training data sets for the TFs and models considered in Figure 8 and Supple- mentary Figures 4 to 13. cell type TF ENCODE-ID GM12878 NRF1 ENCFF652BRY HepG2 ATF3 ENCFF137OEY HepG2 BHLHE40 ENCFF361YXC HepG2 FOXA2 ENCFF184NAC HepG2 HNF4A ENCFF072CXB HepG2 MAX ENCFF140PUO HepG2 YY1 ENCFF177YDT K562 CREM ENCFF021XJN K562 ELF1 ENCFF617ZLL K562 JUND ENCFF213EYD p=1.3e−281 ATF3 (liver vs. K562) − PWM: R=0.178 >:21786 <:2383 =:27954 7.5 7.9e−09 4 −26 5.5e−08 0.083 5.0 ● ● ● 2 ● ● ● ● ● ● −28 2.5 262 5554 368 0 0.0 log(signal)

∆ −30 0.4625 −2 score (K562)

● −2.5 ● ● ● ● ● ● ● ● ● −5.0 ● −4 −32 equal greater less −2 0 2 4 −32 −30 −28 −26 score ∆ score score (liver)

p=0 ATF3 (liver vs. K562) − LSlim: R=0.198 >:27013 <:1959 =:23151 1.2e−14 4 −26 1.1e−14 0.017

5 ● ● 2 ● ● ● ● ● −28 ● ● ● ● ● ● 366 5676 608 0

log(signal) 0 −30 0.634 ∆

−2 score (K562)

● ● ● ● ● ● ● ● ● ● ● ● ● −32 −5 −4 equal greater less −2 0 2 −32 −30 −28 −26 score ∆ score score (liver)

p=0 ATF3 (liver vs. K562) − baseline: R=0.213 >:2283 <:45583 =:4257 7.5 2.7e−37 5.3e−42 0.3 0.63 2.5 5.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● 0.2 1262 0.0 1965 2957 0.0 log(signal) ∆

0.5302 0.1 −2.5 ● −2.5 ●

● (K562) methylation ● ● ● ● ● ● ● ● ● ● ● ● −5.0 ● 0.0 equal greater less −0.2 −0.1 0.0 0.1 0.2 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (liver)

Figure 4: Association between differential model scores and differential binding accord- ing to ChIP-seq data for ATF3 in liver and K562 cell types using PWM models, LSlim models, or a baseline model using average methylation levels in a fixed- size region at the peak. p=5.8e−18 BHLHE40 (K562 vs. GM12878) − PWM: R=0.327 >:8285 <:6553 =:23896 7.5 9e−52 4 4.4e−38 −27 5.0 1.4e−09 ● ● −28 ● 2 ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● 863 −29 12550 483 0 0.0 −30 log(signal) ∆ 0.7791

● ● ● ● ● ● ● ● −2 ● score (GM12878) −31 −2.5 ● ● ● ● ● ● ● ● ● ● ● −5.0 ● −32 equal greater less −2 0 2 −32 −30 −28 score ∆ score score (K562)

p=2.5e−85 BHLHE40 (K562 vs. GM12878) − baseline: R=0.24 >:17522 <:14103 =:7109 7.5 4.4e−150 5.0 3.7e−99 0.3 5.0 0.0038 ● ● ● 2.5 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 2.5 ● ● ● 4296 7252 2348 0.0 0.0 log(signal) ∆ 0.4382 0.1

● ● ● ● ● ● ● ● ● ● ● ● ● ● −2.5 −2.5 ● ● ● ● ● ● ● ● ● ● ● ● ● ●

● (GM12878) methylation −5.0 ● 0.0 equal greater less −0.2 −0.1 0.0 0.1 0.2 0.3 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (K562)

Figure 5: Association between differential model scores and differential binding accord- ing to ChIP-seq data for BHLHE40 in K562 and GM12878 cell types using PWM models or a baseline model using average methylation levels in a fixed- size region at the peak. p=1.1e−207 CREM (K562 vs. GM12878) − PWM: R=0.461 >:10021 <:1533 =:15901 3 6 7.1e−07 1.3e−20 9.3e−19 2 −26

● 4 ● ● ● ● ● ● ● ● ● 1 ● ● ● ● ● ● ● ● ● 2 ● ● −28 13144 262 0 211

log(signal) 0 ∆ score (K562)

0.6436 −1 ● ● ● ● −30 ● ● ● −2 ● ● ● ● ● ● ● ● −2 ● ● −4 equal greater less −2 0 2 −31−30−29−28−27−26 score ∆ score score (GM12878)

p=0 CREM (K562 vs. GM12878) − baseline: R=0.213 >:3362 <:20306 =:3787 6 0.00028 7.1e−43 3.2e−43 2

● 4 ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2 ● ● 0 9269 2571 1777

log(signal) 0 0.1 ∆ 0.3045 ● ● ● ● ● ● ● ● ● ● −2 ● ● −2 ● ● (K562) methylation ● ● ● ● ● ● ● ● ● ● ● ● 0.0 −4 equal greater less −0.2 −0.1 0.0 0.1 0.2 0.3 0.0 0.1 0.2 methylation ∆ methylation methylation (GM12878)

Figure 6: Association between differential model scores and differential binding accord- ing to ChIP-seq data for CREM in K562 and GM12878 cell types using PWM models or a baseline model using average methylation levels in a fixed-size region at the peak. p=3.7e−22 ELF1 (HepG2 vs. GM12878) − PWM: R=0.273 >:4347 <:1154 =:16375 0.00083 2.8e−09 2 5.0 5.9e−09 −26

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● ● ● ● ● ● 0 ● −28 235 11822 242 0.0 log(signal)

∆ −30 0.434 −2 ● ● ● ● ● ● score (GM12878) ● ● ● ● ● ● −2.5 ● ● ● ● −32 ● ● equal greater less −2 0 2 −32 −30 −28 −26 score ∆ score score (HepG2)

p=0 ELF1 (HepG2 vs. GM12878) − baseline: R=0.135 >:5174 <:13169 =:3533 3.6e−54 4.8e−28 5.0 0.97 2 ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● ● ● ● ● ● ● ● ● ● ● ● 1854 0 8429 2029 0.0 log(signal) 0.1 ∆ 0.2275

● −2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● −2.5 ● ● ● ● ● ● ● ● ● (GM12878) methylation ● ● −4 0.0 equal greater less −0.3 −0.2 −0.1 0.0 0.1 0.2 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (HepG2)

Figure 7: Association between differential model scores and differential binding accord- ing to ChIP-seq data for ELF1 in HepG2 and GM12878 cell types using PWM models or a baseline model using average methylation levels in a fixed-size region at the peak. p=0.022 FOXA2 (liver vs. HepG2) − PWM: R=0.305 >:2835 <:857 =:30049 6 2.7e−12 −24 4.1e−11 2 4 0.0037 ● ● ● ● ● 1 ● ● ● −26 2 ● 345 15658 264 0 0 log(signal)

0.526 −28 ∆ −1 ● score (HepG2) ● ● ● ● −2 ● ● ● ● ● ● ● ● ● ● −2 −4 −30 equal greater less −1 0 1 −30 −28 −26 −24 score ∆ score score (liver)

p=0.00033 FOXA2 (liver vs. HepG2) − LSlim: R=0.33 >:3288 <:751 =:29702 6 2.5e−12 2 1.3e−13 −26 4 2e−04 ● ● ● ● 1 ● ● ● ● ● ● 2 ● 294 0 −28 15918 244 0 log(signal)

∆ −1 0.5473

score (HepG2) −30 ● ● ● ● ● −2 ● ● ● ● ● ● −2 ● ● ● ● −4 ● equal greater less −2 −1 0 1 2 −30 −28 −26 score ∆ score score (liver)

p=0 FOXA2 (liver vs. HepG2) − baseline: R=0.239 >:4563 <:24075 =:5103 6 4.7e−58 1.5e−117 4 1.3e−18 ● ● ● 2 ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2 ● ● 6412 4445 5607 0 0 log(signal) 0.1 ∆ 0.3688

● ● ● ● ● −2 ● ● ● ● ● ● ● ● ● −2 ● ● ●

● (HepG2) methylation ● ● ● ● ● ● ● ● ● ● −4 ● ● 0.0 equal greater less −0.2 0.0 0.2 0.0 0.1 0.2 methylation ∆ methylation methylation (liver)

Figure 8: Association between differential model scores and differential binding accord- ing to ChIP-seq data for FOXA2 in HepG2 and liver cell types using PWM models, LSlim models, or a baseline model using average methylation levels in a fixed-size region at the peak. p=0.36 HNF4A (liver vs. HepG2) − PWM: R=0.331 >:3923 <:2969 =:51256 2.8e−08 4 −22 7.3e−17 5.0 2.4e−13 ● ● ● ● 2 −24 ● ● ● ● ● ● ● ● ● 2.5 ● 15601 215 310 0 −26 0.0 log(signal) ∆ 0.681 score (liver)

● ● ● ● ● ● ● ● ● −2.5 ● ● ● −28 ● −2 ● ● ● ● equal greater less −1 0 1 2 −28 −26 −24 −22 score ∆ score score (HepG2)

p=0.057 HNF4A (liver vs. HepG2) − LSlim: R=0.351 >:5297 <:3802 =:49049 7.2e−09 4 −24 2.5e−16 5.0 2.7e−11 ● ● ● 2 −26 ● ● ● ● ● ● ● ● ● 2.5 ● 15840 244 326 0 −28

log(signal) 0.0 ∆ score (liver) 0.6154

● ● ● ● ● ● ● −30 −2.5 ● ● ● −2 ● ● ● ● ● equal greater less −2 −1 0 1 2 3 −30 −28 −26 −24 score ∆ score score (HepG2)

p=6.7e−35 HNF4A (liver vs. HepG2) − baseline: R=0.297 >:26245 <:23699 =:8204 3.4e−80 4 6.7e−172 0.3 5.0 7.5e−42 ● ● ● 2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 2.5 ● ● ● ● 6838 3244 6046 0 0.0 log(signal)

∆ 0.1 0.4991

● ● ● ● ● ● ● ● ● ● (liver) methylation ● ● ● ● ● −2.5 ● ● ● ● −2 ● ● ● ● ● ● ● ● ● 0.0 equal greater less −0.2 −0.1 0.0 0.1 0.2 0.0 0.1 0.2 methylation ∆ methylation methylation (HepG2)

Figure 9: Association between differential model scores and differential binding accord- ing to ChIP-seq data for HNF4A in liver and HepG2 cell types using PWM models, LSlim models, or a baseline model using average methylation levels in a fixed-size region at the peak. p=0.003 JUND (liver vs. HepG2) − PWM: R=0.2 >:3286 <:1865 =:21151 3 6 5.6e−06 −24 6.8e−18 2e−19 2 4 ● ● ● ● ● ● ● 1 ● ● ● ● ● −26 2 ● 249 10395 550 0

log(signal) 0 ∆

−1 score (liver) 0.5373 −28

● ● ● ● −2 ● ● ● −2 ● ● ● ● ● ● ● ● equal greater less −2 −1 0 1 2 3 −28 −26 −24 score ∆ score score (HepG2)

p=1e−07 JUND (liver vs. HepG2) − LSlim: R=0.223 >:5347 <:2731 =:18224 6 5.2e−05 3 1.1e−16 4 4.3e−20 2 −26 ● ● ● ● ● ● ● ● 1 ● ● ● ● 2 ● ● ● 258 −28 11081 545 0

log(signal) 0

∆ −1 score (liver) 0.5121

● ● −30 ● ● ● −2 ● ● −2 ● ● ● ● ● ● ● ● −3 equal greater less −2 0 2 4 −31−30−29−28−27−26 score ∆ score score (HepG2)

p=8.5e−37 JUND (liver vs. HepG2) − baseline: R=0.229 >:10698 <:11615 =:3989 6 1.5e−17 0.25 2.6e−61 1.2e−29 2 4 ● 0.20 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2 ● ● 0.15 1945 0 4899 4350 0.10

log(signal) 0 ∆ 0.3354

● −2 ● ● 0.05 ● ● (liver) methylation ● ● ● ● −2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.00 equal greater less −0.2 0.0 0.2 0.00 0.05 0.10 0.15 0.20 0.25 methylation ∆ methylation methylation (HepG2)

Figure 10: Association between differential model scores and differential binding accord- ing to ChIP-seq data for JUND in liver and HepG2 cell types using PWM models, LSlim models, or a baseline model using average methylation levels in a fixed-size region at the peak. p=6.4e−120 MAX (K562 vs. GM12878) − PWM: R=0.315 >:14374 <:4196 =:27375 9.8e−05 3 3.9e−17 4 −20 5.6e−17 2 ● ● ● ● ● ● −21 ● ● ● ● ● ● ● ● 1 2 ● −22 13872 355 432 0 0 −23 log(signal) ∆ 0.4442

−1 score (K562) −24 ● ● ● ● ● −2 ● ● ● ● ● −2 ● ● ● ● ● −25 ● equal greater less −2 0 2 −24 −22 −20 score ∆ score score (GM12878)

p=3.9e−148 MAX (K562 vs. GM12878) − LSlim: R=0.318 >:16291 <:3749 =:25905 0.00044 3 6.8e−14 4 1.5e−13 2 ● ● ● ● ● −28 ● ● ● ● ● ● ● ● 2 ● ● 1 351 14405 492 0 0 −30 log(signal) ∆ 0.3536

−1 score (K562)

● ● ● ● −2 ● ● ● ● ● −2 ● ● −32 ● ● ● equal greater less −2 0 2 −32 −30 −28 score ∆ score score (GM12878)

p=0 MAX (K562 vs. GM12878) − baseline: R=0.137 >:7482 <:31923 =:6540 1.7e−10 0.3 1.1e−23 4 3.1e−10 2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● 2 ● 9768 2491 2400 0 0 log(signal) 0.1 ∆ 0.1816

● ● ● ● ● ● ● −2 (K562) methylation ● ● ● −2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.0 equal greater less −0.2 −0.1 0.0 0.1 0.2 0.3 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (GM12878)

Figure 11: Association between differential model scores and differential binding accord- ing to ChIP-seq data for MAX in K562 and GM12878 cell types using PWM models, LSlim models, or a baseline model using average methylation levels in a fixed-size region at the peak. p=1.8e−21 NRF1 (K562 vs. HepG2) − PWM: R=0.328 >:4501 <:1674 =:5543 0.00079 4 2e−19 5.2e−23 −26 5 ● 2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● −28 ● ● ● ● ● ● ● ● ● ● ● 9594 753 0 451 −30 log(signal) 0 ∆ score (K562) ● 0.4769 ● ● ● −2 ● ● ● ● ● ● ● ● ● ● −32 ● ● ● ● ● ● ● ● ● ● ● ● −5 ● −4 equal greater less −2 −1 0 1 2 −32 −30 −28 −26 score ∆ score score (HepG2)

p=0 NRF1 (K562 vs. HepG2) − baseline: R=0.205 >:1658 <:7834 =:2226 0.082 4 4.1e−33 1.6e−43 5 ● ● 2 ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0 7160 2203 1442

log(signal) 0 0.1 ∆

● 0.2904 ● ● ● ● ● −2 ● ● ● ● ● ● ● ● ● ● ● ● ● (K562) methylation ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● −5 ● −4 0.0 equal greater less −0.2 0.0 0.2 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (HepG2)

Figure 12: Association between differential model scores and differential binding accord- ing to ChIP-seq data for NRF1 in K562 and HepG2 cell types using PWM models or a baseline model using average methylation levels in a fixed-size region at the peak. p=0.0016 YY1 (K562 vs. GM12878) − PWM: R=0.265 >:3112 <:2378 =:16844 0.00045 6.6e−14 5.0 7.3e−13 2 ● −21 ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● ● ● ● ● ● ● ● ● ● 672 0 −23 15763 350

log(signal) 0.0 ∆ 0.3682 ● ● ● −25 ● ● ● ● ● ● ● −2 score (GM12878) ● ● ● ● ● ● ● ● −2.5 ● ● ● ● ● ● ● equal greater less −2 0 2 −25 −23 −21 score ∆ score score (K562)

p=3e−34 YY1 (K562 vs. GM12878) − baseline: R=0.168 >:10833 <:7610 =:3891 0.56 0.25 2.8e−38 5.0 1.9e−53 2 0.20 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2.5 ● ● 0.15 ● ● ● ● ● ● ● ● ● ● ● 0 10894 3810 2053 0.10

log(signal) 0.0 ∆ 0.2469 ● −2 ● ● ● ● ● ● ● ● ● 0.05 ● ● ● ● ● ● ● ● ● ● ● ● ● ● −2.5 ● ● ● ● ● ● ● (GM12878) methylation ● ● −4 0.00 equal greater less −0.2 0.0 0.2 0.0 0.1 0.2 0.3 methylation ∆ methylation methylation (K562)

Figure 13: Association between differential model scores and differential binding accord- ing to ChIP-seq data for YY1 in K562 and GM12878 cell types using PWM models or a baseline model using average methylation levels in a fixed-size region at the peak. References

1. Xuan Lin, Q. X. et al. (Jan, 2019) MethMotif: an integrative cell speciﬁc database of transcription factor binding motifs coupled with DNA methylation proﬁles. Nucleic Acids Res., 47(D1), D145–D154.

2. Domcke, S., Bardet, A. F., Adrian Ginno, P., Hartl, D., Burger, L., and Sch¨ubeler, D. (2015) Competition between DNA methylation and transcription factors determines binding of NRF1. Nature, 528(7583), 575–579.

3. Yin, Y., Morgunova, E., Jolma, A., Kaasinen, E., Sahu, B., Khund-Sayeed, S., Das, P. K., Kivioja, T., Dave, K., Zhong, F., Nitta, K. R., Taipale, M., Popov, A., Ginno, P. A., Domcke, S., Yan, J., Sch¨ubeler, D., Vinson, C., and Taipale, J. (2017) Impact of cytosine methylation on DNA binding speciﬁcities of human transcription factors. Science, 356(6337).

4. Lea, A. J., Vockley, C. M., Johnston, R. A., Del Carpio, C. A., Barreiro, L. B., Reddy, T. E., and Tung, J. (dec, 2018) Genome-wide quantiﬁcation of the eﬀects of DNA methylation on human gene regulation. eLife, 7, e37513.

5. Kribelbauer, J. F., Laptenko, O., Chen, S., Martini, G. D., Freed-Pastor, W. A., Prives, C., Mann, R. S., and Bussemaker, H. J. (2017) Quantitative Analysis of the DNA Methylation Sensitivity of Transcription Factor Complexes. Cell Reports, 19(11), 2383 – 2395.

6. Zuo, Z. et al. (11, 2017) Measuring quantitative eﬀects of methylation on transcription factor-DNA binding aﬃnity. Sci Adv, 3(11), eaao1799.

7. Mann, I. K., Chatterjee, R., Zhao, J., He, X., Weirauch, M. T., Hughes, T. R., and Vinson, C. (2013) CG methylated microarrays identify a novel methylated sequence bound by the CEBPB—ATF4 heterodimer that is active in vivo. Genome Research, 23(6), 988–997.

8. Suzuki, T., Maeda, S., Furuhata, E., Shimizu, Y., Nishimura, H., Kishima, M., and Suzuki, H. (2017) A screening system to identify transcription factors that induce binding site-directed DNA demethylation. Epigenetics & Chromatin, 10(1), 60.

9. Stadler, M. B., Murr, R., Burger, L., Ivanek, R., Lienert, F., Sch¨oler,A., Nimwegen, E. v., Wirbelauer, C., Oakeley, E. J., Gaidatzis, D., Tiwari, V. K., and Sch¨ubeler, D. (2011) DNA-binding factors shape the mouse methylome at distal regulatory regions. Nature, 480(7378), 490–495.

10. Feldmann, A., Ivanek, R., Murr, R., Gaidatzis, D., Burger, L., and Sch¨ubeler, D. (12, 2013) Transcription Factor Occupancy Can Mediate Active Turnover of DNA Methylation at Regulatory Regions. PLOS Genetics, 9(12), 1–10. 11. Maurano, M. T., Wang, H., John, S., Shafer, A., Canﬁeld, T., Lee, K., and Stama- toyannopoulos, J. A. (2020/04/20, 2015) Role of DNA Methylation in Modulating Transcription Factor Occupancy. Cell Reports, 12(7), 1184–1195.

12. Tian, H.-P., Lun, S.-M., Huang, H.-J., He, R., Kong, P.-Z., Wang, Q.-S., Li, X.-Q., and Feng, Y.-M. (2015) DNA Methylation Aﬀects the SP1-regulated Transcription of FOXF2 in Breast Cancer Cells. Journal of Biological Chemistry, 290(31), 19173– 19183.

13. Harrington, M. A., Jones, P. A., Imagawa, M., and Karin, M. (1988) Cytosine methylation does not aﬀect binding of transcription factor Sp1. Proceedings of the National Academy of Sciences, 85(7), 2066–2070.

14. H¨oller,M., Westin, G., Jiricny, J., and Schaﬀner, W. (1988) Sp1 transcription factor binds DNA and activates transcription even when the binding site is CpG methylated.. Genes & Development, 2(9), 1127–1135.

15. Kim, J. D., Hinz, A. K., Choo, J. H., Stubbs, L., and Kim, J. (2007) YY1 as a controlling factor for the Peg3 and Gnas imprinted domains. Genomics, 89(2), 262 – 269.

16. Blattler, A., Yao, L., Wang, Y., Ye, Z., Jin, V. X., and Farnham, P. J. (2013) ZBTB33 binds unmethylated regions of the genome associated with actively ex- pressed genes. Epigenetics & Chromatin, 6(1), 13.

17. Prokhortchouk, A., Hendrich, B., Jørgensen, H., Ruzov, A., Wilm, M., Georgiev, G., Bird, A., and Prokhortchouk, E. (2001) The p120 catenin partner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes & Development, 15(13), 1613–1618.

18. Daniel, J. M., Spring, C. M., Crawford, H. C., Reynolds, A. B., and Baig, A. (07, 2002) The p120 ctn-binding partner Kaiso is a bi-modal DNA-binding protein that recognizes both a sequence-speciﬁc consensus and methylated CpG dinucleotides . Nucleic Acids Research, 30(13), 2911–2919.