TABLE 1.
Summary of modeling strategies and performances
| Target | Difficulty | Interchain contact | Modeling method | Model 1 | Best out of 5 (based on Z‐score) | ||||
|---|---|---|---|---|---|---|---|---|---|
| Z‐score a | ICS | TM‐score (oligo) | Z‐score b | ICS | TM‐score (oligo) | ||||
| H1036 c | Medium | No | Template | 0.84 | 0.68 | 0.70 | 1.17 | 0.72 | 0.71 |
| H1036v0 c | Medium | No | Template | 0.93 | 0.27 | 0.69 | 0.92 | 0.27 | 0.69 |
| H1045 | Medium | No | Template | 1.05 | 0.71 | 0.87 | 1.35 | 0.77 | 0.87 |
| H1047 | Hard | Yes (G) | ab initio | 1.35 | 0.04 | 0.39 | 1.54 | 0.04 | 0.38 |
| H1060v1 | Medium | No | ab initio | 0.94 | 0.06 | 0.31 | 1.09 | 0.08 | 0.31 |
| H1060v2 | Medium | No | Template | 0.38 | 0.09 | 0.86 | 0.34 | 0.10 | 0.88 |
| H1060v3 | Medium | No | Template | 0.43 | 0.01 | 0.75 | 0.85 | 0.12 | 0.84 |
| H1060v4 | Medium | No | Template | 0.74 | 0.22 | 0.75 | 0.93 | 0.21 | 0.73 |
| H1060v5 | Medium | Yes (G) | Template | 1.52 | 0.48 | 0.95 | 1.67 | 0.50 | 0.95 |
| H1065 | Hard | Yes (DL) | Fold‐and‐dock | 1.74 | 0.40 | 0.79 | 1.82 | 0.40 | 0.79 |
| H1072 | Medium | Yes (DL) | Fold‐and‐dock | 0.10 | 0.04 | 0.34 | 0.28 | 0.03 | 0.37 |
| H1081v0 | Medium | No | ab initio | 1.46 | 0.35 | 0.97 | 1.59 | 0.35 | 0.97 |
| H1097 | Medium | Yes (T) | Fold‐and‐dock | 2.13 | 0.44 | 0.73 | 1.99 | 0.44 | 0.74 |
| T1032 | Easy | No | Template | 1.08 | 0.38 | 0.69 | 1.10 | 0.40 | 0.68 |
| T1034 | Medium | No | Template | −0.81 | 0.00 | 0.17 | −0.38 | 0.00 | 0.23 |
| T1038 | Hard | No | ab initio | −0.58 | 0.00 | 0.17 | 0.36 | 0.01 | 0.20 |
| T1048 | Medium | Yes (DL) | Fold‐and‐dock | 3.09 | 0.50 | 0.59 | 4.29 | 0.58 | 0.83 |
| T1052 | Easy | No | Template | 0.63 | 0.51 | 0.69 | 0.72 | 0.51 | 0.69 |
| T1054 | Hard | No | ab initio | 0.13 | 0.00 | 0.44 | 0.81 | 0.00 | 0.52 |
| T1061 | Hard | No | Template | 1.85 | 0.15 | 0.64 | 1.97 | 0.17 | 0.69 |
| T1070 | Hard | No | Template | 0.94 | 0.06 | 0.31 | 2.10 | 0.10 | 0.37 |
| T1078 | Medium | No | ab initio | 0.19 | 0.00 | 0.54 | 2.50 | 0.25 | 0.67 |
| T1080 | Hard | Yes (DL) | Fold‐and‐dock | 1.92 | 0.12 | 0.55 | 2.60 | 0.13 | 0.61 |
| T1083 | Medium | Yes (DL) | Fold‐and‐dock | 1.48 | 0.23 | 0.63 | 1.60 | 0.23 | 0.63 |
| T1084 | Medium | Yes (DL) | Fold‐and‐dock | 2.17 | 0.81 | 0.92 | 2.20 | 0.84 | 0.91 |
| T1087 | Medium | Yes (DL) | Fold‐and‐dock | 2.27 | 0.36 | 0.79 | 2.86 | 0.36 | 0.79 |
| T1099v0 | Medium | No | Template | −0.23 | 0.03 | 0.24 | 0.22 | 0.02 | 0.45 |
| T1099v1 | Medium | No | Template | 0.15 | 0.00 | 0.55 | 0.27 | 0.00 | 0.55 |
| T1099v2 | Medium | No | Template | 0.75 | 0.13 | 0.60 | 0.89 | 0.16 | 0.60 |
Abbreviations: DL, Deep learning‐based methods; G, GREMLIN; T, Partial templates.
a
Calculated on model 1 submissions.
b
Calculated on all model submissions.
c
Having a completely wrong prediction for the antigen‐antibody interface.