Skip to main content
Fig. 2 | Journal of Cheminformatics

Fig. 2

From: Detecting drug communities and predicting comprehensive drug–drug interactions via balance regularized semi-nonnegative matrix factorization

Fig. 2

Predicting comprehensive interactions in the cold-start scenario. a A toy example of predicting how likely the new drug ‘X’ interacts with other drugs in terms of pharmacological changes. In the network shown in the left panel, the nodes numbered from 1 to 7 are drugs having known comprehensive interactions and the isolated node labeled by ‘X’ denotes the newly given drug having no interactions with the previous drugs. Blue single lines and yellow double lines denote enhancive (positive) and degressive (negative) interactions respectively. The adjacent matrix of the DDI network is shown the right panel. b The framework of the BRSNMF-based predictive approach. All the dimensions of the matrices are also listed

Back to article page