squeakus / grn Goto Github PK

Thanks for the great code. Helped me a lot to code my version of GRNs. I found an issue with your code and although I'm not sure how this whole issue reporting thing works, I figured it might be helpful to share it here.

On line 170 of the grn.py, the same max_observed value has been calculated for both enh_matrix and inh_matrix which shouldn't be the case judging from the Banzhaf's GRN formula. I believe you did this code based on Nicolau's later paper which doesn't include separate formulas for both enhancers and inhibitors. So I believe your code should be changed to:

enh_matrix = enh_matrix - np.max(enh_matrix)
inh_matrix = inh_matrix - np.max(inh_matrix)