Table 2

Performance of SVMHC for different HLA alleles, using MHCPEP- or SYFPEITHI-data.
MHC
 Length
 Size
 Mc
 Kernel
Predictions using MHCPEP
HLA-A1
 9
 28
 0.95
 lin
HLA-A*1101
 9
 40
 0.74
 poly
HLA-A11
 9
 46
 0.75
 rbf
HLA-A11
 10
 21
 0.59
 poly
HLA-A2
 9
 118
 0.76
 poly
HLA-A2
 10
 35
 0.65
 poly
HLA-A*2402
 9
 73
 0.90
 poly
HLA-A3
 9
 73
 0.76
 rbf
HLA-A*0201
 9
 184
 0.73
 rbf
HLA-A*0201
 10
 96
 0.78
 poly
HLA-A*3301
 9
 32
 0.72
 lin
HLA-A*0301
 9
 38
 0.72
 rbf
HLA-A*0301
 10
 32
 0.77
 lin
HLA-A31
 9
 39
 0.79
 poly
HLA-A*6801
 9
 42
 0.84
 poly
HLA-B7
 9
 32
 0.95
 lin
HLA-B8
 9
 26
 0.77
 poly
HLA-B*2705
 9
 41
 0.93
 lin
HLA-B*3501
 9
 67
 0.93
 lin
HLA-B*3501
 10
 34
 0.96
 poly
HLA-B35
 9
 23
 0.71
 lin
HLA-B*2703
 9
 22
 0.90
 lin
HLA-B*5301
 9
 41
 0.95
 lin
HLA-B27
 9
 34
 0.91
 rbf
HLA-B*2706
 9
 20
 0.93
 lin
HLA-B51
 9
 67
 0.82
 poly
HLA-B*5102
 9
 29
 0.79
 poly
HLA-B*0702
 9
 52
 0.96
 poly
HLA-B*5103
 9
 29
 0.84
 rbf
HLA-B*5401
 9
 42
 0.98
 lin
HLA-B*5101
 9
 35
 0.89
 lin
Predictions using SYFPEITHI
HLA-A*0201
 9
 113
 0.78
 rbf
HLA-A*0201
 10
 40
 0.70
 poly
HLA-A1
 9
 28
 0.96
 lin
HLA-A3
 9
 73
 0.80
 lin
HLA-B*8
 8
 14
 0.89
 lin
HLA-B*8
 9
 25
 0.79
 lin
HLA-B*2705
 9
 29
 1.00
 lin
HLA-B7
 9
 23
 0.93
 lin

The first column explains shows the HLA allele, the second the length of the binding peptides, the third the number of binders included in the training set, the fourth the performance as measured by the Matthews correlation coefficient. The final column shows what type of kernel was used in the Support Vector Machine.

Dönnes and Elofsson BMC Bioinformatics 2002 3:25   doi:10.1186/1471-2105-3-25