Fig. 9
From: CEMiTool: a Bioconductor package for performing comprehensive modular co-expression analyses
Hypothetical R2 x β curves for two hypothetical studies. a A good choice for β parameter since the corresponding R2 value is above a specified threshold (indicated by the blue line), and subsequent β values brings no additional significant increase in adherence to the scale-free model (b) A poor choice for β parameter since even though it’s the smallest possible beta value above the given threshold, it presents significantly less adherence to the scale-free model than subsequent values of beta