Table S3. Predictive performance (AUC) of 58 algorithmic combinations in the training and three independent validation cohorts ID Training GSE49541 GSE25097 GSE36411 Lasso+SVM 0.949989117 0.873454102 0.979166667 0.87952381 glmBoost+RF 0.947042566 0.872081046 0.979166667 0.895396825 glmBoost 0.943888355 0.896650756 0.95 0.895396825 RF+Lasso 0.937406257 0.807595359 0.820833333 0.872721088 glmBoost+Lasso 0.933825924 0.811515035 0.8125 0.868185941 glmBoost+Enet[alpha=0.5] 0.930516463 0.833111589 0.795833333 0.859115646 glmBoost+Enet[alpha=0.9] 0.930247102 0.813596869 0.804166667 0.863650794 RF+Enet[alpha=0.6] 0.928284219 0.841330772 0.7875 0.840975057 Enet[alpha=0.6] 0.925931316 0.8617751 0.716666667 0.890861678 RF+NaiveBayes 0.920747032 0.823811924 0.7375 0.902199546 Lasso 0.91847604 0.791748361 0.8125 0.856848073 RF+Enet[alpha=0.5] 0.917315057 0.820419817 0.7875 0.852312925 RF+glmBoost 0.915409629 0.612738088 0.845833333 0.899931973 RF+Enet[alpha=0.7] 0.915059914 0.812655341 0.791666667 0.852312925 glmBoost+Enet[alpha=0.8] 0.91487916 0.793031151 0.804166667 0.856848073 RF+Enet[alpha=0.1] 0.906539177 0.811164417 0.779166667 0.86138322 RF+Enet[alpha=0.8] 0.906098307 0.79802449 0.795833333 0.854580499 Enet[alpha=0.7] 0.904347396 0.84715357 0.725 0.874988662 glmBoost+plsRglm 0.901856158 0.723248482 0.8625 0.850045351 plsRglm 0.897988713 0.743844806 0.76875 0.906734694 Lasso+Stepglm[forward] 0.889375194 0.848190269 0.797916667 0.766145125 RF+plsRglm 0.888867732 0.713743618 0.81875 0.878390023 Enet[alpha=0.5] 0.88716269 0.835854634 0.691666667 0.881791383 glmBoost+Stepglm[forward] 0.886163526 0.782868493 0.80625 0.81829932 Enet[alpha=0.9] 0.879460887 0.744607434 0.791666667 0.870453515 Lasso+glmBoost 0.868237456 0.558471599 0.95 0.895396825 Enet[alpha=0.4] 0.856815704 0.804193428 0.695833333 0.893129252 Lasso+NaiveBayes 0.852331462 0.690166456 0.789583333 0.886326531 RF+Enet[alpha=0.9] 0.850814777 0.698205332 0.704166667 0.863650794 Lasso+plsRglm 0.846261749 0.698343 0.81875 0.848911565 glmBoost+NaiveBayes 0.840249397 0.678855022 0.795833333 0.890861678 glmBoost+Enet[alpha=0.7] 0.837288945 0.697313012 0.8 0.865918367 glmBoost+Enet[alpha=0.3] 0.831667493 0.714746617 0.783333333 0.863650794 Enet[alpha=0.8] 0.820954051 0.726850899 0.7375 0.872721088 RF+Enet[alpha=0.4] 0.819054346 0.7002534 0.783333333 0.852312925 glmBoost+Enet[alpha=0.6] 0.815329749 0.675132141 0.795833333 0.859115646 RF+LDA 0.812448532 0.687658223 0.779166667 0.856848073 Enet[alpha=0.3] 0.801687651 0.721198972 0.695833333 0.90446712 Enet[alpha=0.2] 0.793090543 0.719753663 0.691666667 0.909002268 RF+Ridge 0.792441836 0.646252535 0.758333333 0.913537415 RF+Enet[alpha=0.2] 0.786575588 0.678826612 0.783333333 0.854580499 Enet[alpha=0.1] 0.776253163 0.696875553 0.7 0.918072562 RF+Enet[alpha=0.3] 0.773823639 0.676273459 0.783333333 0.854580499 Lasso+LDA 0.773205886 0.679064714 0.783333333 0.847777778 glmBoost+LDA 0.770466111 0.697351295 0.741666667 0.863650794 NaiveBayes 0.766986844 0.740821472 0.70625 0.854580499 Ridge 0.760908778 0.686051843 0.704166667 0.911269841 glmBoost+Ridge 0.757187225 0.624039466 0.725 0.893129252 RF+Stepglm[forward] 0.741693483 0.741705818 0.747916667 0.752539683 glmBoost+Enet[alpha=0.2] 0.732034927 0.601640854 0.770833333 0.86138322 glmBoost+Enet[alpha=0.4] 0.731660933 0.591057595 0.7875 0.854580499 glmBoost+Enet[alpha=0.1] 0.730274085 0.591029674 0.754166667 0.872721088 Stepglm[forward] 0.723217825 0.808649867 0.633333333 0.75707483 glmBoost+SVM 0.716109699 0.743153552 0.770833333 0.665238095 RF+SVM 0.715879321 0.714470742 0.6625 0.736666667 LDA 0.713523811 0.818996901 0.55 0.727596372 Lasso+RF 0.708439262 0.647702578 0.6875 0.665238095 SVM 0.701745001 0.610247746 0.6875 0.641428571