consensus subnetworks results

[water wu]

Hello,

When i run the paper/paper_commands.sh (hotnet2 v1.2.1) with the default settings for  example data, the consensus folder contains four files. The stats table shows

#Size	Observed	Expected	P-value
2	14	0	1
3	10	0	1
4	9	0	1
5	9	0	1
6	6	0	1
7	5	0	1
8	3	0	1
9	3	0	1
10	3	0	1

Is the result (Expected  and P-value) of consensus correct? or do i always need to use -cp argument in HotNet2.py? 

I get the similar result (Expected=0 and P-value=1 for all sizes) with my own data in consensus mode.

(The results on individual networks in both data don't have these issue.)

Thanks for your help.

[matthe...@brown.edu]

Hello,

With its default settings, HotNet2 uses zero permutations to evaluate the statistical significance of the HotNet2 consensus subnetworks, so you should disregard the results of the "Expected" and "P-value" entries for the consensus subnetworks when using the default settings.  To evaluate the statistical significance of the HotNet2 consensus subnetworks, please set the -cp (--consensus permutations) argument for the HotNet2.py script, e.g., -cp 100.

Best,

Matt