Parameter Estimation, Graph Alignment
Parameter estimation is a process of finding an appropriate set of parameters for the scoring function of the graph alignment problem ( Scoring Function, Graph Alignment). These parameters serve the same purpose as the scoring matrices of the Sequence Alignment (e.g., PAM or BLOSUM ( BLOcks SUbstitution Matrix (BLOSUM))). The scoring parameters are estimated for both the node score ( Node Score, Graph Alignment) and the link score ( Link Score, Graph Alignment) and they ultimately weigh the relative contributions of topological and sequence similarity to the score ( Scoring Function, Graph Alignment).
The early parameter estimates were determined ad hoc and allowed only limited application of the graph alignment ( Graph Alignment, Protein Interaction Networks) (e.g., motif search ( Network Motif), pathway querying ( Pathway), and an alignment based solely on the protein similarity). With an increased availability of the...