Event:

Complex phenotypes are determined by many genes and can well be investigated using network methods. Hence, to study the evolution of complex phenotypes, comparative network approaches are very promising. Here we aim to shed light on the evolution of cognitive abilities in humans. To this end, we used transcriptome data from pre-frontal cortex samples of humans, chimpanzees, and rhesus macaques to calculate transcription factor (TF) co-expression networks. We revealed an overall increase in connectivity on the human lineage and that several TFs that are known to be involved in brain development or cognitive disorders have turned into hubs specifically in the human networks. These TFs are good candidates for further experimental studies on the evolution of cognition. We developed several methods to facilitate comparative network analyses, among them a statistical framework to combine similar networks into consensus networks of high confidence, to assign p-values to links, and to define conserved,species-specific, and diverged links. Our methods are publicly available via two R packages, wTO and CoDiNA, and come along with sophisticated interactive visualization. We hope them to be useful for other evolutionary network studies.