A bug in function netVisual_chord_gene

[shenscore]

Hi,

I found a bug lately and I realized that this will misleading users: When the names of ligand and receptor of an interaction are same (such as CDH2_CDH2 and MPZL1_MPZL1), this will cause a big problem when try to plot chord diagram using chordDiagram function.

Here are two plots before and after I rewrite this function. The original netVisual_chord_gene function generated a disordered chord plot.
I'm afraid that other similar functions may have the same problem.