Similarity indices

PyNetSim could calculate five similarity groups composed of 23 similarity measures, including 13 common neighbor-based similarity indices, one preferential attachment index, 3 network path-based similarity indices, 5 random walk-based similarity indices, and one matrix forest index. These similarity indices capture and characterize different node relationships by applying different scientific theories to the network under investigation. We hope that the similarity indices calculated by this package could be useful when solving various biomedical and drug research questions in the context of network biology, network pharmacology and network medicine.

Index type Index name Abbreviation
Common neighbor-based similarity Common neighbors CN >>Enter
Salton index Salton >>Enter
Jaccard index Jaccard >>Enter
Sorensen index Sorensen >>Enter
Hub promoted index HPI >>Enter
Hub depressed index HDI >>Enter
Leicht-Holme-Newman Index LHN1 >>Enter
Adamic-Adar index AA >>Enter
Resource allocation index RA >>Enter
Geometric index Geo >>Enter
First Kulczynski index FK >>Enter
Second Kulczynski index SK >>Enter
Sokal and Sneath UN2 index SSUN2 >>Enter
Preferential attachment score Preferential attachment index PA >>Enter
Network path-based similarity Local path index LP >>Enter
Katz index Katz >>Enter
Leicht-Holme-Newman index LHN2 >>Enter
Random walk-based similarity Average commute time ACT >>Enter
Cosine based on L+ CosPlus >>Enter
SimRank SimRank >>Enter
Local random walk LRW >>Enter
Superposed random walk SRW >>Enter
Matrix forest index Matrix forest index MFI >>Enter