The first sentence of the second paragraph in the Introduction makes the broad statement that existing cancer-specific pathways analysis methods do not take into account normal tissue expression, however, this is not generally true. In particular, the Pathifier method (reference 12 in the paper) does consider expression in the matching normal tissue when computing pathway dysregulation scores. For interested readers, the primary difference between the Pathifier and TPAC methods lies in how the pathway dysregulation scores are computed. Specifically, the Pathifier technique calculates pathway dysregulation scores using the distance between the projection of each tumor sample and the associated normal tissue samples along a principal curve computed in a low dimensional subspace of the pathway expression data. By comparison, the TPAC method quantifies pathway dysregulation using the gamma CDF value for a modified Mahalanobis distance computed between the pathway-specific transcriptomic profile for each tumor and a reference normal tissue profile. Features unique to TPAC include the use of the normal tissue-specificity of each pathway gene to adjust scores and support for single sample inference.

My sincere thanks to Proferror Eytan Domany for noticing the oversight!