RNA

Supplementary Data

HDOCK: a web server for protein-protein and protein-DNA/RNA docking based on a hybrid strategy Yumeng Yan, Di Zhang, Pei Zhou, Botong Li, and Sheng-You Huang ∗

School of Physics, Huazhong University of Science and Technology, Wuhan Hubei 430074, P. R. China

Key Words: docking web server, protein-protein docking, protein-DNA docking, protein-RNA docking, template-based docking.

∗

Email: [email protected]; Phone: +86-27-87543881; Fax: +86-027-87556576

1

Table S1: Protein-protein docking results of free unbound docking, HDOCK server, and templatebased modeling for the 54 cases with weakly homologous complex templates from the docking benchmark by the Weng group. Target PDB ID 1AK4 1AVX 1BKD 1BUH 1CGI 1EAW 1F6M 1FC2 1FFW 1FQ1 1GCQ 1GHQ 1GL1 1GPW 1H9D 1HE1 1IBR 1IRA 1J2J 1JK9 1KAC 1KTZ 1MAH 1N8O 1OPH 1PVH 1QA9 1R8S 1S1Q 1SBB 1TMQ 1WQ1 1XD3 1XQS 1Y64 1YVB 1Z5Y 1ZLI 2A5T 2ABZ 2AYO 2CFH 2G77 2HRK 2I25 2J0T 2J7P 2NZ8 2OUL 2OZA 2UUY 3CPH 3D5S 3SGQ Successf

Free unbound dockinga Accuracyb Rankc 3 131 2 1 3 82 2 10 3 62 2 1 3 56 3 12 3 41 – – 2 19 3 2622 3 5 3 1 3 315 3 228 – – – – 3 3 2 7 3 226 – – 1 2 2 1 3 13 2 1667 3 1564 – – 2 29 2 1668 2 172 3 77 3 5 3 588 – – 3 4 3 112 – – 3 767 3 80 3 9 2 1 3 4 3 38 3 267 3 260 3 126 3 33 1 1 3 880 3 17 3 17 2 66 1 4 11.1% (29.6%)

Templated 2XGY:B/A 3E8L:A/C 3CF6:E/R 3GQI:A/B 1CT4:E/I 1TOC:B/R 3QFA:A/C 4IOI:H/A 1SRR:A/A 2B4S:A/B 4OMN:A/A 2XWB:B/F 3MYW:A/I 2ISS:A/D 3QDJ:E/A 1G4U:S/R 1QBK:C/B 4DEP:C/A 2ODE:A/B 1QUP:A/B 2J1K:C/J 1LX5:A/B 4QWW:A/B 4IW4:E/C 3H5C:A/B 2Q7N:A/B 1CDC:A/B 3TNF:A/B 1UZX:A/B 4RGN:E/A 4CVW:A/C 5CJP:E/A 4WLR:A/C 3QML:C/A 4EAH:A/D 3KSE:A/D 1VRS:D/A 4BD9:A/B 4PE5:A/C 1ZLH:A/B 5L8W:A/C 2J3W:D/F 4HLQ:A/B 5BMU:B/A 4N1E:A/I 3CKI:A/B 5L3Q:A/B 1FOE:A/B 3CBJ:A/B 3GCU:A/B 2FTM:A/B 1VG9:A/B 2WY7:A/Q 1PPF:E/I

a Standard

HDOCK server Accuracyb Rankc 3 15 3 74 2 1 2 11 2 1 3 147 3 528 – – 3 31 – – – – 3 1097 3 26 3 14 3 42 2 1 – – – – 3 10 2 123 3 7 3 1468 – – 2 1 3 5 3 1618 – – 3 1071 3 12 3 904 2 27 3 250 2 21 – – 3 1 2 6 2 2 – – 2 2137 3 4 2 1 3 1 2 1 2 8 3 232 1 1 2 43 2 1 2 1 – – 2 1 2 1 3 245 3 3 24.1% (38.9%)

Modeling/Accuracye – – 2 – 3 – – – – – – – – – – 2 – – – 2 – – – – – – – – 2 – – – 2 – 3 2 2 – – 3 2 3 3 2 – 3 2 2 2 – 2 2 – 2 38.9%

free unbound docking with the unbound structures from the benchmark using HDOCKlite. accuracy of the first correct solution, represented by 1 (high), 2 (medium), 3 (acceptable), and “–” (incorrect), respectively. c The rank of the first correct solution with at least acceptable accuracy. “–” means no correct solution found. d The weakly homologous complex template used for template-based docking with HDOCK server. e The accuracy of the template-based model predicted by homology modeling with the weakly complex template. f The success rates for free unbound docking, HDOCK, and template-based modeling when considering the top 1 (top 10) prediction(s). b The

2

Table S2: Protein-DNA docking results of free unbound docking, HDOCK server, and template-based modeling for the 23 cases with weakly homologous complex templates from the docking benchmark by the Bonvin group. Target PDB ID 1BDT 1CMA 1DDN 1DFM 1DIZ 1EMH 1G9Z 1JT0 1RPE 1RXW 1VRR 1Z9C 1ZS4 1ZTW 2HDD 2IRF 2ODI 2R1J 3BAM 3CRO 3JXY 4KTQ 7MHT Successf

Free unbound dockinga Accuracyb Rankc 3 275 3 97 3 326 – – 3 439 3 867 – – 3 30 2 284 2 19 3 1587 3 206 3 208 2 289 3 1 3 13 – – 2 134 3 2382 2 59 3 2 3 262 3 15 4.3% (8.7%)

Templated 2O4A:A/B 2IH2:A/B 3H0D:A/D 1VRR:A/C 3I0W:A/B 4YIG:A/C 2QOJ:Z/X 3V21:A/I 6CRO:A/U 5T9J:A/D 3ODH:A/C 4LLN:A/G 1B7F:A/P 4C8Z:A/C 1O4X:A/C 1PUE:E/A 1MW8:X/Y 3O9X:A/E 1VRR:A/C 4PU3:C/P 2I3Q:A/C 1SL1:A/P 4DKJ:A/B

a Standard

HDOCK server Accuracyb Rankc – – 3 163 3 754 2 1 3 114 3 23 3 41 3 1343 3 1 3 57 3 73 2 147 – – 3 192 3 4 3 20 – – 3 610 3 244 3 380 2 5 3 1 2 1 17.4% (26.1%)

Modeling/Accuracye – – – 2 – 3 – – 3 – – – – – 3 3 – – – – – – 3 26.1%

free unbound docking with the unbound structures from the benchmark using HDOCKlite. accuracy of the first correct solution, represented by 1 (high), 2 (medium), 3 (acceptable), and “–” (incorrect), respectively. c The rank of the first correct solution with at least acceptable accuracy. “–” means no correct solution found. d The weakly homologous complex template used for template-based docking with HDOCK server. e The accuracy of the template-based model predicted by homology modeling with the weakly complex template, represented by 1 (high), 2 (medium), 3 (acceptable), and “–” (incorrect). f The success rates for free unbound docking, HDOCK, and template-based modeling when considering the top 1 (top 10) prediction(s). b The

3

Table S3: Protein-RNA docking results of free unbound docking, HDOCK server, and template-based modeling for the 33 cases with weakly homologous complex templates from the docking benchmark by the Fernandez group. Target PDB ID 1B7F 1C0A 1DFU 1EFW 1F7U 1FEU 1FXL 1H3E 1HQ1 1K1G 1K8W 1M5O 1M8W 1N78 1QF6 1QTQ 1RKJ 1U0B 1VFG 1WSU 2AD9 2BTE 2BU1 2CSX 2DB3 2ERR 2FY1 2IX1 2NUG 2PY9 3BSO 3BSX 3BX2 Successf

Free unbound dockinga Accuracyb Rankc 2 1 3 157 3 4 2 11 2 27 3 102 3 1 – – 3 1 3 3 2 1 – – 2 8 1 1 – – – – – – – – 3 77 3 884 3 1 3 3256 3 4 3 16 3 470 2 31 3 2 2 1 – – 3 491 3 879 3 6 3 10 21.2% (42.4%)

Templated 3NMR:A/B 1ASY:A/R 1FEU:A/C 3KFU:A/L 2ZUE:A/B 1D6K:A/B 2QFJ:A/C 2DLC:X/Y 3ADD:A/C 1B23:P/R 1ZE2:A/C 3NNC:A/B 3V71:A/B 1EUY:A/B 1H4S:A/T 1N78:A/C 4YB1:P/R 5AH5:A/C 3WFR:G/C 2PJP:A/B 1M5O:C/A 1WZ2:A/C 4ANG:A/R 1WZ2:A/C 2NMV:A/D 4WKR:A/C 4N0T:A/B 2VNU:D/B 4OOG:C/D 5U4J:c/x 4K50:A/C 3K5Q:A/B 3V71:A/B

a Standard

HDOCK server Accuracyb Rankc 3 28 2 1 3 33 2 1 2 1 3 1 3 5 3 113 3 17 3 5 2 1 3 27 2 1 2 1 – – 3 97 – – 3 72 3 1 3 21 3 46 3 1 2 82 2 3 1 2 2 3 3 3 2 4 3 13 3 25 3 142 2 1 2 1 33.3% (54.5%)

Modeling/Accuracye – – – 2 2 3 – – – – 2 – 2 – – – – – – 2 – – – – – – – – – – – 2 2 24.2%


4

Table S4: Protein-RNA docking results of free unbound docking, HDOCK server, and template-based modeling for the 33 cases with weakly homologous complex templates from the docking benchmark by the Bahadur group. Target PDB ID 1ASY 1C0A 1CVJ 1DFU 1F7U 1FEU 1G2E 1K8W 1M5O 1M8W 1MMS 1N78 1QF6 1QTQ 1VFG 1WSU 2BTE 2GIC 2GXB 2HW8 2IX1 2XNR 2XZO 3AEV 3BSN 3BX2 3NMR 3O3I 3QJJ 4J7M 4YCP 4ZT0 5F5H Successf

Free unbound dockinga Accuracyb Rankc – – 2 47 3 3 3 7 2 27 2 19 3 1 2 1 – – 2 7 3 18 1 1 3 594 – – 2 7 3 20 3 120 3 18 2 8 – – 2 2 3 3 3 36 1 1 3 1842 1 4 2 17 3 444 2 1 2 3 1 1 – – 1 18 18.2% (45.5%)

Templated 1IL2:A/C 1ASY:A/R 2J0S:D/E 1FEU:A/C 2ZUE:A/B 1D6K:A/B 2QFJ:A/C 1ZE2:A/C 3NMR:A/B 3V71:A/B 2VJU:A/C 1EUY:A/B 1H4S:A/T 1N78:A/C 3WFR:G/C 2PJP:A/B 1WZ2:A/C 2GTT:A/W 4KMF:A/B 1MZP:A/B 4PMW:A/C 2XS2:A/B 4B3G:A/G 2ANR:A/B 3OL9:A/C 3V71:A/B 1B7F:A/P 1J4W:A/B 3PKM:A/G 5BUD:A/D 3B0V:C/A 5CZZ:A/B 4NDH:A/D

a Standard

HDOCK server Accuracyb Rankc 3 12 2 1 1 1 2 3 2 1 3 173 3 94 2 1 – – 3 1 3 86 2 1 2 8 3 667 3 10 3 45 2 35 3 14 2 9 2 1631 2 1 3 1 3 51 1 1 3 4 3 1 3 4 2 109 2 1 3 29 3 85 – – 2 11 33.3% (51.5%)

Modeling/Accuracye 3 2 – 3 2 3 – 2 – 2 – – – – – 2 – – – – – 3 – – – 3 3 – 2 – – – – 36.4%


5

Table S5: Protein-RNA docking results of free unbound docking, HDOCK server, and template-based modeling for the 25 cases with weakly homologous complex templates from the docking benchmark by Huang and Zou. Target PDB ID 1C0A 1DFU 1F7U 1FFY 1GAX 1H3E 1IL2 1J1U 1JID 1K8W 1N78 1QTQ 1R3E 1T0K 1U0B 2AKE 2BTE 2CSX 2RFK 2UWM 2ZM5 2ZUE 3EPH 3FOZ 3HHZ Successf

Free unbound dockinga Accuracyb Rankc 2 1 2 2 3 38 1 1 1 1 3 62 3 1335 3 466 3 78 2 1 2 1 1 1 1 1 2 51 3 1274 2 5 3 46 1 1 2 1 3 348 1 1 1 1 1 1 1 1 3 9 52% (64%)

Templated 1ASY:A/R 1FEU:A/C 2ZUE:A/B 1GAX:A/C 1FFY:A/T 2DLC:X/Y 1ASY:A/R 1ASY:A/R 3KTW:A/C 1ZE2:A/C 1EUY:A/B 1N78:A/C 1K8W:A/B 3PLA:C/G 5AH5:A/C 1J1U:A/B 1WZ2:A/C 1WZ2:A/C 1K8W:A/B 2PJP:A/B 3EPH:A/E 1WZ2:A/C 2ZM5:A/C 3EPH:A/E 2GTT:A/W

a Standard

HDOCK server Accuracyb Rankc 2 1 2 1 2 1 3 7 3 1 3 341 2 1 3 271 3 561 2 1 2 1 2 1 2 12 3 19 3 497 3 240 3 1 3 2 3 65 3 29 2 1 1 1 2 1 2 1 3 2 52% (64%)

Modeling/Accuracye 2 2 2 – – – 3 – – 2 – 3 – – – – – – – 2 2 – 2 2 – 40%


6

Table S6: Comparison of HDOCK and template-based modeling for the 20 targets of CAPRI round 30 for which we have identified a template for modeling an oligomer of subunits. The results were taken from Ref. 23 in the main manuscript. CAPRI Target T69 T71 T72 T73 T75 T78 T79 T80 T81 T82 T84 T85 T86 T87 T89 T90 T91 T92 T93 T94 Success

Templatea 1QLW 4H09 3MX3 3H3L 3RJT 1Z84 3RCO 3BN1 2W83 4DN2 1M7O 2G5C 3FF2 3GET 1SHY 1MDO 1BYR 1GLP 3G79 3GFF

fnat 0.094 0 0 0.025 0.586 0 0.04 0.746 0 0.498 0.737 0.615 0 0.763 0.475 0.558 0 0.594 0.59 0.019

Template-based modeling ˚ ˚ Lrmsd (A) Irmsd (A) Accu.b 43.734 18.679 – 40.671 14.082 – 53.374 21.243 – 46.357 17.495 – 43.803 2.363 ** 26.433 18.135 – 19.664 11.71 – 4.367 1.997 ** 22.362 6.565 – 34.722 4.309 * 4.930 2.247 ** 3.214 2.598 ** 48.628 17.765 – 2.669 1.949 ** 12.199 4.26 * 6.781 3.974 ** 15.406 17.337 – 8.346 4.508 ** 6.319 3.191 ** 46.637 19.364 – 50%

a

fnat 0.407 0.45 0.37 0 0.534 0.25 0.16 0.535 0.509 0.339 0.473 0.67 0 0.603 0.66 0.548 0.573 0.301 0.542 0.329

HDOCK ˚ ˚ Lrmsd (A) Irmsd (A) 3.458 3.165 17.331 2.689 13.316 4.546 7.427 3.337 3.629 1.88 21.682 7.731 17.917 11.101 3.828 1.837 5.077 2.425 3.17 2.649 3.391 2.062 1.924 1.933 39.649 11.516 2.66 2.28 3.885 2.412 3.197 1.887 5.189 1.877 3.484 3.896 3.718 2.563 8.3 4.701

Accu.b ** * – – ** – – ** * ** ** ** – ** ** ** ** ** ** * 75%

The PDB code of the complex template used for template-based docking to predict the assembly of subunits. The accuracy is categorized by three parameters following the CAPRI criteria: the percentage of the native residueresidue contacts (f nat ), the ligand RMSD (L rmsd ), and the interface RMSD (I rmsd ). “***” stands for high-accuracy, “**” for medium-accuracy, “*” for acceptable accuracy, and “–” for no correct prediction, respectively.

b

7