Table 4.

Confusion matrices from the PAM algorithm for class prediction of NCI-60 tissue of origin

BRCNSCOLCLEMEOVREClass error rate
Protein
    BR112000010.8
    CNS060000000
    CO006000100.14
    LC001610100.33
    LE000060000
    ME010009000.1
    OV010000510.29
    RE000000080
Consensus
    BR022100001
    CNS060000000
    CO007000000
    LC011400210.56
    LE020040000.33
    ME010009000.1
    OV002100400.43
    RE000100160.25
HU133A
    BR022100001
    CNS060000000
    CO007000000
    LC011400210.56
    LE000050010.17
    ME010009000.1
    OV001500100.86
    RE000200060.25
HU95A
    BR022100001
    CNS060000000
    CO007000000
    LC021300030.67
    LE010041000.33
    ME010009000.1
    OV003000220.71
    RE000100160.25
HU6800
    BR022100001
    CNS050000010.17
    CO006100000.14
    LC021200220.78
    LE010050000.17
    ME010009000.1
    OV003400001
    RE012000140.5
cDNA
    BR012100011
    CNS050000010.17
    CO006000100.14
    LC020310210.67
    LE100050000.17
    ME010009000.1
    OV001300300.57
    RE000100160.25
  • NOTE: Rows indicate “true” classifications; columns indicate “predicted” classifications.

  • Abbreviations: BR, breast cancer; CO, colon; LC, lung cancer; LE, leukemia; ME, melanoma; OV, ovarian; RE, renal.