Skip to main content

Table 3 Model Performance

From: A systematic analysis of genomics-based modeling approaches for prediction of drug response to cytotoxic chemotherapies

  NLSVR PCR LNSVR ANN
NFS DEG LIM NFS DEG LIM NFS DEG LIM NFS DEG LIM
BLM .207 .202 .202 .239 .208 .208 .151 .1 .209 .147 .17 .21
BTZ .38 .404 .365 .422 .399 .354 .332 .326 .232 −.009 .299 .24
CIS .05 .08 N/A −.009 .047 N/A .03 .079 N/A −.066 .034 N/A
CYT .313 .32 .279 .32 .281 .256 .337 .291 .269 .226 .266 .291
DTX .422 .44 .408 .367 .409 .382 .357 .319 .359 .185 .318 .207
DOX .273 .27 .117 .243 .285 .106 .27 .226 .103 .115 .173 .096
ETP .289 .302 .294 .248 .291 .263 .238 .219 .273 .209 .195 .246
GEM .143 .139 .166 .153 .117 .143 .07 .063 .165 .131 .119 .134
MTX .461 .455 .462 .431 .435 .433 .417 .388 .338 .411 .391 .322
MMC .237 .302 .244 .264 .269 .25 .27 .224 .239 .203 .153 .248
PTX .32 .27 .198 .287 .282 .159 .233 .170 .191 −.106 .211 .177
VBL .44 .403 .399 .408 .398 .37 .398 .339 .371 .112 .302 .363
VOR .509 .495 .486 .5 .487 .439 .484 .471 .404 .445 .42 .42
SN-38 .383 .417 .409 .379 .391 .443 .397 .404 .429 .01 .327 .402
5-FU .463 .464 .40 .455 .484 .354 .451 .438 .337 .309 .409 .365
AVG .326 .331 .316 .314 .319 .297 .285 .27 .28 .144 .252 0.266
  1. Average spearman correlations across six different testing sets for all regression and feature selection methods. This data is graphically displayed in Fig. 3