Fig. 2

Schematic description of the proposed method. The proposed method consists of a total 4 steps: a it builds two disease networks PrDN and DrDN using protein and drug information respectively. b this step selects candidate disease by prioritizing diseases whose difference in edges is very high by mirroring DrDN from PrDN c it scores on other diseases against candidate disease through machine learning algorithm and then selects diseases whose score is high as similar diseases and then assigns candidate drugs which is used for such diseases d lastly, it repositions candidate drugs onto candidate disease