Table 1 The comparison between the existing studies is presented
From: SAlign–a structure aware method for global PPI network alignment
Method | Features | Topological method | Alignment heuristic | Datasets | Advantages | Limitations |
|---|---|---|---|---|---|---|
HubAlign | Sequence + topology | Min. degree heuristic | Greedy algorithm | IntAct | Scalable better alignment in terms of no. of aligned nodes | AFS is not better as HubAlign prioritises topology |
ModuleAlign | Sequence + topology + clustering based scores | Min. degree heuristic + cluster similarity scores | Hungarian algorithm | HINT | Module based (clustering) scoring matrix helps in producing quality alignment | Complexity is high |
PROPER | Sequence + topology | Local network topology | Percolation graph matching algorithm | IntAct | Takes less resources and time | Align few no. of nodes |
IBNAL | Functional similarity + topology | Clique-degree signature similarity | Greedy Algorithm (based on clique size) | IsoBase | Uses less resources | Go-annotations are required in alignment phase |
MAGNA | Sequence only | – | Genetic algorithm | BioGRID | Efficient for alignments that required high topological quality | 1-optimize the results w.r.t topology only that results in low semantic similarity 2-exponential complexity time |
NETAL | Topology only | Local topological measure with iterative updates | Greedy-algorithm | IntAct | High speed | Performance is measured using topological measures only |
UAlign | Sequence + topology | UAlign unifies the alignments of eight aligners which include Natalie, SPINAL, PISwap, MAGNA, HubAlign, L-GRAAL, OptNetAlign and ModuleAlign. The best features of all aligners are used to optimize the alignment w.r.t different measures | ||||