Is there a way to determine a threshold (p-value, FDR...) for significant contacts?

Thank you,
Gorliver