Multi-level learning: improving the prediction of protein, domain and residue interactions by allowing information flow between levels

Table 1 Data features at the protein level.

Feature	Feature of	Data type	Kernel
COG (version 7) phylogenetic profiles [62]	Proteins	Binary vectors	Linear
Sub-cellular localization [22]	Proteins	Binary vectors	Linear
Cell cycle gene expression [24]	Proteins	Real vectors	Correlation (linear after standardization)
Environment response gene expression [63]	Proteins	Real vectors	Correlation (linear after standardization)
Yeast two-hybrid [1, 4]	Protein pairs	Unweighted graph	Diffusion (β = 1)
TAP-MS [5, 6]	Protein pairs	Weighted graph	Diffusion (β = 1)

ISSN: 1471-2105