Table 2.
Table 2.1. Optimized s value for rotamer libraries of size ranging from 0.2Å to 5.0Å, plus the all torsion rotamer library. The s values for that gives the best RMSD value is listed. | ||||
---|---|---|---|---|
Library | Number of Rotamers |
Unmodified Potential (RMSD,Å) |
Best s value | Best RMSD (Å) |
0.2Å | 14755 | 0.536 | 1.3 | 0.468 |
0.6Å | 3195 | 0.710 | 1.1 | 0.564 |
1.0Å | 1014 | 0.857 | 1.2 | 0.665 |
1.4Å | 378 | 0.958 | 1.1 | 0.753 |
1.8Å | 214 | 1.064 | 0.9 | 0.885 |
2.2Å | 136 | 1.343 | 0.8 | 1.175 |
3.0Å | 84 | 1.624 | 0.7 | 1.487 |
5.0Å | 44 | 1.890 | 0.7 | 1.860 |
All Torsion | 382 | 0.937 | 0.9 | 0.800 |
Table 2.2. Effect of s values on χ1/χ1+2 accuracy. Rotamer libraries of diversity ranging from 0.2Å to 5.0Å, plus the all torsion rotamer library are used. The best χ1+2 accuracy is used to determine the most effective scaling factor c. A χ angle is considered correct if within 40° of the corresponding χ angle in the crystal sidechain conformation. | ||||
---|---|---|---|---|
Library | Number of Rotamers |
χ1/χ1+2 accuracy from unmodified scoring function |
Best scaling factor s |
χ1/χ1+2 accuracy using best s value |
0.2Å | 14755 | 95.0% / 91.8% | 1.3 | 96.3% / 93.4% |
0.6Å | 3195 | 92.6% / 87.7% | 1.1 | 95.6% / 92.1% |
1.0Å | 1014 | 90.0% / 83.4% | 1.2 | 95.3% / 90.4% |
1.4Å | 378 | 87.8% / 80.0% | 1.2 | 94.7% / 88.9% |
1.8Å | 214 | 84.3% / 75.6% | 1.2 | 91.5% / 83.8% |
2.2Å | 136 | 71.9% / 61.0% | 0.8 | 79.1% / 68.0% |
3.0Å | 84 | 63.4% / 54.1% | 0.7 | 68.4% / 58.9% |
5.0Å | 44 | 53.2% / 44.9% | 0.7 | 54.9% / 45.8% |
All Torsion | 382 | 89.6% / 81.3% | 1.1 | 93.3% / 86.8% |