Fig. 2

CTpathway algorithm diagram including Multi-RWR. First, gene 1 in the expression profile was taken out as a seed and RWR was used to obtain the crosstalk effect C₁ on all nodes in the network. Next, another gene i was chosen to repeat this progress and obtained C_i. Finally, we obtained the crosstalk effect matrix C after all genes in the profile were taken out as a seed. We also calculated the differential expression (DE) score by integrating FC and P-value. Using both crosstalk and differential expression, we obtained the risk score \({RS}_{j_i}\) of gene j impacted by gene i. We integrated the risk score of gene j impacted by all genes as gene j risk score RS_j. Finally, we obtained a pathway risk score (PS) by averaging all gene risk scores in a pathway and calculated the significance level by permutation