Skip to main content

Table 3 Results on MKL Experiments on 5 × 5 cross-validation experiments

From: A multiple kernel learning algorithm for drug-target interaction prediction

Dataset Combination Pairs Targets Drugs
NR [SPEC-k4]-[AERS-freq] 0.4630 (±0.0215) 0.3851 (±0.0254) 0.2341 (±0.0054)
  [SPEC-k4]-[GIP] 0.5187 (±0.0255) 0.3725 (±0.0247) 0.0949 (±0.0068)
  BLM-KA 0.0709 (±0.0048) 0.3441 (±0.0264) 0.3130 (±0.0224)
  BLM-MEAN 0.0685 (±0.0062) 0.3453 (±0.0264) 0.2934 (±0.0154)
  KBMF2MKL 0.2041 (±0.0150) 0.2059 (±0.0388) 0.1459 (±0.0272)
  KRONRLS-KA 0.4321 (±0.0147) 0.3489 (±0.0337) 0.2850 (±0.0126)
  KRONRLS-MEAN 0.4078 (±0.0211) 0.3482 (±0.0341) 0.2665 (±0.0109)
  KRONRLS-MKL 0.5368 (±0.0137) 0.3541 (±0.0321) 0.3383 (±0.0224)
  LAPRLS-KA 0.1989 (±0.0207) 0.2120 (±0.0277) 0.1841 (±0.0044)
  LAPRLS-MEAN 0.1870 (±0.0196) 0.2008 (±0.0251) 0.1832 (±0.0022)
  NETLAPRLS-KA 0.2310 (±0.0277) 0.2091 (±0.0288) 0.1841 (±0.0044)
  NETLAPRLS-MEAN 0.2195 (±0.0273) 0.1989 (±0.0263) 0.1831 (±0.0023)
  NRWRH-KA 0.1776 (±0.0380) 0.1911 (±0.0116)
  NRWRH-MEAN 0.1755 (±0.0364) 0.1881 (±0.0109)
  PKM-KA 0.1830 (±0.0114) 0.2363 (±0.0387) 0.1741 (±0.0158)
  PKM-MAX 0.0946 (±0.0188) 0.0774 (±0.0108) 0.1174 (±0.0080)
  PKM-MEAN 0.1702 (±0.0099) 0.2163 (±0.0400) 0.1672 (±0.0152)
  SITAR 0.4477 (±0.0658) 0.1396 (±0.0505) 0.0694 (±0.0189)
  WANG-MKL 0.3293 (±0.0175) 0.2238 (±0.0300) 0.2628 (±0.0225)
GPCR [SPEC-k4]-[MINMAX] 0.3246 (±0.0093) 0.5053 (±0.0322) 0.0924 (±0.0055)
  [SW]-[GIP] 0.6188 (±0.0075) 0.4561 (±0.0201) 0.0419 (±0.0014)
  BLM-KA 0.0633 (±0.0071) 0.5508 (±0.0123) 0.3000 (±0.0198)
  BLM-MEAN 0.0519 (±0.0032) 0.5353 (±0.0135) 0.2526 (±0.0188)
  KBMF2MKL 0.4960 (±0.0124) 0.0963 (±0.0346) 0.1408 (±0.0120)
  KRONRLS-KA 0.6208 (±0.0081) 0.4727 (±0.0101) 0.3005 (±0.0148)
  KRONRLS-MEAN 0.6213 (±0.0085) 0.4461 (±0.0086) 0.2731 (±0.0155)
  KRONRLS-MKL 0.6440 (±0.0052) 0.4127 (±0.0126) 0.3161 (±0.0112)
  LAPRLS-KA 0.2183 (±0.0067) 0.1458 (±0.0050) 0.1210 (±0.0058)
  LAPRLS-MEAN 0.2169 (±0.0066) 0.1369 (±0.0049) 0.1215 (±0.0061)
  NETLAPRLS-KA 0.3763 (±0.0096) 0.1451 (±0.0041) 0.1211 (±0.0062)
  NETLAPRLS-MEAN 0.3841 (±0.0088) 0.1357 (±0.0039) 0.1221 (±0.0061)
  NRWRH-KA 0.0762 (±0.0041) 0.1201 (±0.0088)
  NRWRH-MEAN 0.0704 (±0.0036) 0.1176 (±0.0099)
  PKM-KA 0.2625 (±0.0133) 0.2327 (±0.0175) 0.1424 (±0.0146)
  PKM-MAX 0.1230 (±0.0106) 0.0652 (±0.0071) 0.0935 (±0.0044)
  PKM-MEAN 0.2613 (±0.0178) 0.1632 (±0.0186) 0.1254 (±0.0107)
  SITAR 0.5324 (±0.0267) 0.1151 (±0.0538) 0.0283 (±0.0110)
  WANG-MKL 0.4240 (±0.0071) 0.3521 (±0.0111) 0.2686 (±0.0274)
IC [PPI]-[GIP] 0.6789 (±0.0078) 0.1548 (±0.0020) 0.0467 (±0.0009)
  [SW]-[GIP] 0.8679 (±0.0056) 0.7301 (±0.0140) 0.0476 (±0.0008)
  BLM-KA 0.1169 (±0.0127) 0.7944 (±0.0047) 0.2516 (±0.0304)
  BLM-MEAN 0.1106 (±0.0088) 0.7798 (±0.0040) 0.2152 (±0.0257)
  KBMF2MKL 0.7671 (±0.0033) 0.4420 (±0.0141) 0.0856 (±0.0044)
  KRONRLS-KA 0.8553 (±0.0017) 0.7246 (±0.0071) 0.2039 (±0.0190)
  KRONRLS-MEAN 0.8693 (±0.0011) 0.6885 (±0.0067) 0.1887 (±0.0186)
  KRONRLS-MKL 0.8769 (±0.0011) 0.6894 (±0.0056) 0.2406 (±0.0259)
  LAPRLS-KA 0.3088 (±0.0021) 0.2747 (±0.0031) 0.0942 (±0.0022)
  LAPRLS-MEAN 0.3187 (±0.0024) 0.2760 (±0.0032) 0.0939 (±0.0021)
  NETLAPRLS-KA 0.5359 (±0.0065) 0.2750 (±0.0032) 0.0931 (±0.0022)
  NETLAPRLS-MEAN 0.5560 (±0.0073) 0.2766 (±0.0034) 0.0928 (±0.0023)
  NRWRH-KA 0.2371 (±0.0046) 0.0720 (±0.0026)
  NRWRH-MEAN 0.2363 (±0.0042) 0.0712 (±0.0024)
  PKM-KA 0.5133 (±0.0235) 0.4151 (±0.0092) 0.1156 (±0.0041)
  PKM-MAX 0.1608 (±0.0132) 0.1673 (±0.0038) 0.0660 (±0.0031)
  PKM-MEAN 0.5474 (±0.0261) 0.3840 (±0.0062) 0.0998 (±0.0019)
  SITAR 0.7505 (±0.0153) 0.1717 (±0.0633) 0.0174 (±0.0046)
  WANG-MKL 0.7116 (±0.0214) 0.6009 (±0.0158) 0.2217 (±0.0124)
E [GO]-[GIP] 0.6900 (±0.0032) 0.2371 (± 0.0025) 0.0124 (±0.0004)
  [SW]-[GIP] 0.8429 (±0.00540) 0.7438 (± 0.0189) 0.0159 (±0.0003)
  BLM-KA 0.0471 (±0.0045) 0.8201 (±0.0070) 0.2506 (±0.0060)
  BLM-MEAN 0.0374 (±0.0032) 0.8099 (±0.0063) 0.2079 (±0.0051)
  KBMF2MKL 0.6722 (±0.0051) 0.0757 (±0.0049) 0.0213 (±0.0004)
  KRONRLS-KA 0.8630 (±0.0127) 0.7274 (±0.0071) 0.1829 (±0.0034)
  KRONRLS-MEAN 0.8667 (±0.0098) 0.6917 (±0.0062) 0.1655 (±0.0030)
  KRONRLS-MKL 0.8818 (±0.0128) 0.7384 (±0.0063) 0.2168 (±0.0050)
  LAPRLS-KA 0.1920 (±0.0014) 0.1677 (±0.0072) 0.0682 (±0.0012)
  LAPRLS-MEAN 0.1750 (±0.0015) 0.1402 (±0.0055) 0.0646 (±0.0013)
  NETLAPRLS-KA 0.2853 (±0.0024) 0.1669 (±0.0042) 0.0670 (±0.0018)
  NETLAPRLS-MEAN 0.2548 (±0.0019) 0.1402 (±0.0046) 0.0636 (±0.0016)
  NRWRH-KA 0.0886 (±0.0011) 0.0403 (±0.0024)
  NRWRH-MEAN 0.0816 (±0.0006) 0.0383 (±0.0018)
  PKM-KA 0.2383 (±0.0069) 0.1905 (±0.0047) 0.0480 (±0.0037)
  PKM-MAX 0.0762 (±0.0011) 0.0597 (±0.0007) 0.0323 (±0.0007)
  PKM-MEAN 0.2161 (±0.0072) 0.1239 (±0.0032) 0.0382 (±0.0031)
  SITAR 0.7558 (±0.0160) 0.0232 (±0.0151) 0.0097 (±0.0111)
  WANG-MKL 0.7286 (±0.0046) 0.6663 (±0.0069) 0.1648 (±0.0042)
  1. Best performing methods are indicated in bold. Standart deviation is indicated in brackets. Training of the PKM, SITAR and WANG algorithms was done with the balanced training set
  2. best on training
  3. best on testing