Skip to main content
. 2024 May 2;40(5):btae301. doi: 10.1093/bioinformatics/btae301

Table 1.

EmbedPVP variant prediction results across several ontologies with different neuro-symbolic knowledge embedding methods.

Using the clinical phenotypes
Using OMIM phenotypes
H@1 H@10 H@30 H@50 ROCAUC AUPR H@1 H@10 H@30 H@50 ROCAUC AUPR
Genotype-based prediction tools CADD 116 (0.0759) 266 (0.1741) 467 (0.3056) 591 (0.3868) 0.9778 0.0494 116 (0.0759) 266 (0.1741) 467 (0.3056) 591 (0.3868) 0.9778 0.0494
MCAP 4 (0.0026) 261 (0.1708) 442 (0.2893) 511 (0.3344) 0.6389 0.0076 4 (0.0026) 261 (0.1708) 442 (0.2893) 511 (0.3344) 0.6389 0.0076
SIFT 201 (0.1315) 201 (0.1315) 201 (0.1315) 201 (0.1315) 0.6436 0.0736 201 (0.1315) 201 (0.1315) 201 (0.1315) 201 (0.1315) 0.6436 0.0736
PolyPhen2 127 (0.0831) 127 (0.0831) 127 (0.0831) 226 (0.1479) 0.6465 0.0481 127 (0.0831) 127 (0.0831) 127 (0.0831) 226 (0.1479) 0.6465 0.0481
DANN 21 (0.0137) 263 (0.1721) 263 (0.1721) 263 (0.1721) 0.8422 0.0115 21 (0.0137) 263 (0.1721) 263 (0.1721) 263 (0.1721) 0.8422 0.0115
MetaSVM 20 (0.0131) 111 (0.0726) 318 (0.2081) 406 (0.2657) 0.6510 0.0108 20 (0.0131) 111 (0.0726) 318 (0.2081) 406 (0.2657) 0.6510 0.0108
Phenotype-based prediction tools PHIVE 181 (0.1185) 325 (0.2127) 364 (0.2382) 380 (0.2487) 0.8047 0.0709 346 (0.2264) 496 (0.3246) 518 (0.3390) 523 (0.3423) 0.8151 0.1477
DeepPVP 221 (0.1446) 661 (0.4326) 762 (0.4987) 795 (0.5203) 0.7662 0.1389 449 (0.2938) 858 (0.5615) 905 (0.5923) 924 (0.6047) 0.8041 0.2853
Phenix 472 (0.3089) 628 (0.4110) 746 (0.4882) 788 (0.5157) 0.8148 0.2154 1104 (0.7225) 1130 (0.7395) 1153 (0.7546) 1159 (0.7585) 0.8206 0.6275
hiPHIVE 431 (0.2821) 653 (0.4274) 768 (0.5026) 809 (0.5295) 0.8098 0.1982 868 (0.5681) 1025 (0.6708) 1149 (0.7520) 1184 (0.7749) 0.8151 0.4693
EmbedPVP (TransD) GO 307 (0.2009) 563 (0.3685) 726 (0.4751) 829 (0.5425) 0.9524 0.1386 670 (0.4385) 894 (0.5851) 1006 (0.6584) 1042 (0.6819) 0.9795 0.3464
HP 482 (0.3154) 846 (0.5537) 1007 (0.659) 1056 (0.6911) 0.9895 0.2507 996 (0.6518) 1230 (0.805) 1352 (0.8848) 1391 (0.9103) 0.9960 0.5865
MP 396 (0.2592) 675 (0.4418) 868 (0.5681) 947 (0.6198) 0.9587 0.1869 779 (0.5098) 922 (0.6034) 1031 (0.6747) 1072 (0.7016) 0.9822 0.4120
UBERON 287 (0.1878) 509 (0.3331) 674 (0.4411) 800 (0.5236) 0.9493 0.1278 699 (0.4575) 892 (0.5838) 995 (0.6512) 1023 (0.6695) 0.9775 0.3594
Union 409 (0.2677) 639 (0.4182) 833 (0.5452) 928 (0.6073) 0.9581 0.1934 899 (0.5884) 1086 (0.7107) 1158 (0.7579) 1245 (0.8148) 0.9933 0.5087
EmbedPVP (DL2Vec) GO 152 (0.0995) 382 (0.2500) 554 (0.3626) 614 (0.4018) 0.9282 0.0659 491 (0.3213) 804 (0.5262) 944 (0.6178) 1010 (0.6610) 0.9787 0.2485
HP 362 (0.2369) 666 (0.4359) 787 (0.5151) 826 (0.5406) 0.9867 0.1758 1011 (0.6616) 1300 (0.8508) 1366 (0.8940) 1384 (0.9058) 0.9942 0.6168
MP 255 (0.1669) 491 (0.3213) 639 (0.4182) 701 (0.4588) 0.9501 0.1128 639 (0.4182) 914 (0.5982) 1043 (0.6826) 1106 (0.7238) 0.9804 0.3386
UBERON 174 (0.1139) 390 (0.2552) 498 (0.3259) 556 (0.3639) 0.8928 0.0751 539 (0.3527) 801 (0.5242) 904 (0.5916) 940 (0.6152) 0.9271 0.2713
Union 358 (0.2343) 636 (0.4162) 771 (0.5046) 824 (0.5393) 0.9605 0.1673 950 (0.6217) 1216 (0.7958) 1310 (0.8573) 1353 (0.8855) 0.9936 0.5625
EmbedPVP (OWL2Vec*) GO 188 (0.1230) 385 (0.2520) 525 (0.3436) 592 (0.3874) 0.9190 0.0797 557 (0.3645) 876 (0.5733) 1011 (0.6616) 1059 (0.6931) 0.9780 0.2935
HP 409 (0.2677) 685 (0.4483) 783 (0.5124) 842 (0.5510) 0.9874 0.1987 1026 (0.6715) 1313 (0.8593) 1373 (0.8986) 1391 (0.9103) 0.9940 0.6304
MP 222 (0.1453) 470 (0.3076) 618 (0.4045) 677 (0.4431) 0.9508 0.0992 665 (0.4352) 965 (0.6315) 1068 (0.6990) 1116 (0.7304) 0.9785 0.3582
UBERON 158 (0.1034) 379 (0.2480) 474 (0.3102) 525 (0.3436) 0.8866 0.0673 577 (0.3776) 800 (0.5236) 888 (0.5812) 937 (0.6132) 0.9291 0.2937
Union 375 (0.2454) 650 (0.4254) 787 (0.5151) 835 (0.5465) 0.9563 0.1774 959 (0.6276) 1253 (0.8200) 1325 (0.8671) 1368 (0.8953) 0.9939 0.5775

Bold values indicate the highest scores achieved among different models.