Table 5 Different types of string kernels (see Table 1 for further details)

Spectrum Kernel

generates all possible subsequences of length k and counts the occurrences of these k-mers

Density: -

Information: -

[112]

Mismatch Kernel

considers a certain distance, hence mismatches, between two k-mers

Density: -

Information: o

[114,115,116]

General Structure

Distant Segment Kernel

allows a gap between two k-mers

Density: -

Information: o

[118]

Local Alignment Kernel

obtained from local alignment scores

Density: +

Information: o

[119]

Spectrum Kernel, Mismatch Subsequence Kernel

Subsequence Kernel

measures sequence similarity, gaps within k-mers are taken into account

Density: +

Information: o

[119]

Frequency of Amino Acid Pairs

Frequency of Amino Acid Pairs

similar to dipeptide composition

Density: -

Information: o

[119]

String Kernels + Physicochemical Properties

optimization of existing string kernels such that these involve physicochemical properties

Density: +

Information: +

[120]

Physicochemical Properties

Generic String Kernel

string kernel with physicochemical properties and penalization of non adjacent segments

Density: +

Information: +

[121, 122]