Fig. 1

Workflow for the identification of common signatures between blood and brain tissue. The blood and brain datasets are transformed into gene co-expression networks using Weighted Gene Co-expression Network Analysis (WGCNA). The networks are represented by modules, groups of highly co-expressed genes. Modules significantly correlated with HD are identified. The modules are annotated using Concept Profile Analysis (CPA), using four semantic annotation categories i.e. Biological Processes, Cellular Component, Molecular Function and Disease or Syndrome. Next, the similarity between the annotated modules from each network is assessed (for each semantic annotation category) based on the total number of overlapping annotations between each blood-brain module pair (pairwise matching). The significance of each module pair is assessed by repeating the entire analysis using randomly composed modules of the same gene size as the original ones. The random distribution was used to assign a significance value for each module pair. At the bottom of the figure, the module pairs that cross our threshold (P v a l u e<0.05) which compose our two blood-brain signatures