In an effort to mitigate the strong off-target effects of shRNAs, we developed an algorithm to produce a Consensus Gene Signature (CGS) that reflects the consistent (and therefore on-target) gene expression effects of shRNAs. To generate a consensus gene signature (CGS), we first create a pairwise Spearman correlation matrix between all shRNA signatures targeting the same gene, explicitly setting self-correlations to 0. Each shRNA signature is then assigned a weight given by the sum of its correlations to the other signatures, with the weights normalized to sum to 1. The CGS is computed as the linear combination of the shRNA signatures, with coefficients set to the weights.
