cna with explicit cluster.method

Greetings, I want to understand how communities in a residue interaction network are calculated. I have tried two commands that I thought should generate identical output but they do not. I want to understand why:

Run 1: If I run:

my_net8 <- cna(cij.traj, cutoff.cij = cutoff, vnames=res_names_WT)

print.cna(my_net8)

Call: cna.dccm(cij = cij.traj, cutoff.cij = cutoff, vnames = res_names_WT)

Structure: - NETWORK NODES#: 105 EDGES#: 546 - COMMUNITY NODES#: 11 EDGES#: 28 + attr: network, communities, community.network, community.cij, cij, call

Run 2: But, if I run (adding explicitly options for cluster.method, collapse.method, and minus.log):

my_net8 <- cna(cij.traj, cutoff.cij = cutoff, vnames=res_names_WT, cluster.method = "btwn", collapse.method = "max", minus.log = TRUE)

print.cna(my_net8)

Call: cna.dccm(cij = cij.traj.rabbit, cutoff.cij = cutoff, vnames = res_names_WT, cluster.method = "btwn", collapse.method = "max", minus.log = TRUE) Structure: - NETWORK NODES#: 105 EDGES#: 231 - COMMUNITY NODES#: 11 EDGES#: 9 + attr: network, communities, community.network, community.cij, cij, call

The output from Run 1 and Run 2 is different: see Edges#.

When I do "Run 1" after doing "Run 2", then the two outputs are identical. What may be the issue here?

Thanks!

Comments (2)