I know that motif detection is usually done by comparing the number of motifs of a particular type in a real world network again a random network generated. If a significant more number of that particular motif is present in the real world network when compared to the random, then that motif is said be significantly enriched in the network. But what do you say about the opposite.
In my bipartite, directed network, out of the two possible three node motifs (SIM and DOR), I found that SIM was significantly enriched when compared to the corresponding random networks. However, there were significantly lesser DOR motifs in the network when compared to the corresponding random network. Is there a conclusion I can draw from this? Can you guide me to any papers that has reported such a finding.
My reasoning is that having a lesser number of motifs in the real network when compared to the random essentially means that it is not significant and no conclusions can be drawn from that. Any advice is kindly appreciated.
