Fig. 1.

The workflow of the study. Seven proximity measures were used to calculate the shortest path length between the protein targets of a drug and the lncRNA-correlated proteins in the interactome. Performance of the drug indication prediction was compared among different proximity measures. An unsupervised learning method was used to cluster drugs on the basis of seven proximity measures. The proximal drugs were validated by the correlation of the response of the cell lines and the lncRNA gene expression.