Skip to main content

Table 2 Coverage of labeled structures, number of predicted affinities for unlabeled structures, as well as specificity, precision, and accuracy for HPC-based prediction of binding affinity as recovered from SDPs computed using our algorithm (with λ=6 and δ=16). Sensitivity is equal to 1 in all cases

From: Structure-guided selection of specificity determining positions in the human Kinome

Inhibitor Cov. #pred. Spec. Prec. Acc.
ABT-869 86 % 557 0.922 0.633 0.931
AMG-706 83 % 558 0.928 0.707 0.938
AST-487 65 % 426 0.661 0.806 0.859
AZD-1152HQPA 85 % 568 0.914 0.668 0.927
BIRB-796 67 % 391 0.766 0.653 0.838
BMS-387032/SNS-032 96 % 670 0.984 0.959 0.988
CHIR-258/TKI-258 81 % 420 0.947 0.861 0.960
CHIR-265/RAF265 87 % 473 0.960 0.801 0.966
CI-1033 77 % 475 0.882 0.710 0.909
CP-690550 96 % 629 0.989 0.736 0.989
CP-724714 99 % 684 0.999 0.982 0.999
Dasatinib 83 % 500 0.897 0.837 0.933
EKB-569 70 % 474 0.876 0.688 0.902
Erlotinib 80 % 532 0.902 0.693 0.920
Flavopiridol 80 % 515 0.844 0.754 0.895
GW-2580 99 % 677 1.000 1.000 1.000
GW-786034 79 % 485 0.920 0.737 0.934
Gefitinib 81 % 470 0.906 0.561 0.916
Imatinib 86 % 587 0.936 0.590 0.941
JNJ-7706621 59 % 356 0.580 0.704 0.790
LY-333531 83 % 413 0.912 0.652 0.924
Lapatinib 99 % 684 0.999 0.982 0.999
MLN-518 94 % 659 0.989 0.808 0.989
MLN-8054 87 % 493 0.948 0.766 0.956
PI-103 99 % 654 0.999 0.988 0.999
PKC-412 54 % 217 0.621 0.687 0.793
PTK-787 97 % 664 0.999 0.974 0.999
Roscovitine/CYC202 98 % 650 1.000 1.000 1.000
SB-202190 84 % 500 0.929 0.815 0.946
SB-203580 69 % 349 0.792 0.641 0.849
SB-431542 100 % 670 1.000 1.000 1.000
SU-14813 71 % 343 0.761 0.667 0.838
Sorafenib 70 % 509 0.919 0.801 0.939
Staurosporine 91 % 646 0.681 0.956 0.959
Sunitinib 61 % 343 0.652 0.654 0.790
VX-680/MK-0457 78 % 410 0.844 0.767 0.897
VX-745 85 % 583 0.912 0.680 0.926
ZD-6474 87 % 511 0.939 0.823 0.952
average 83 % 520 0.887 0.783 0.929
  1. The last row lists the average performance over all inhibitors