Backup in gene regulatory networks explains differences between binding and knockout results

The complementarity of gene expression and protein–DNA interaction data led to several successful models of biological systems. However, recent studies in multiple species raise doubts about the relationship between these two datasets. These studies show that the overwhelming majority of genes bound by a particular transcription factor (TF) are not affected when that factor is knocked out. Here, we show that this surprising result can be partially explained by considering the broader cellular context in which TFs operate. Factors whose functions are not backed up by redundant paralogs show a fourfold increase in the agreement between their bound targets and the expression levels of those targets. In addition, we show that incorporating protein interaction networks provides physical explanations for knockout effects. New double knockout experiments support our conclusions. Our results highlight the robustness provided by redundant TFs and indicate that in the context of diverse cellular systems, binding is still largely functional. We show that backup in regulatory networks partially explains why the overwhelming majority of genes bound by a particular transcription factor are not affected when that factor is knocked out. Factors whose functions are backed up by redundant paralogs show a fourfold decrease in the agreement between their bound targets and the expressions levels of those targets. Protein interaction networks explain why many of the knockout targets are not physically bound by the knocked out factor. Double knockout expression experiments show that when both the factor and its paralog are knocked out, the agreement between binding and knockout data increases significantly, supporting our conclusions.