hotnet2 consensus permutations

[Armand Bankhead]

Hello,

  I'm trying to evaluate consensus network significance using permutations (via the -np parameter) and am getting the following error:

Traceback (most recent call last):

  File "/home/bankhead/topBfx/bankhead/software/hotnet2/20180626//HotNet2.py", line 139, in <module>

    run(get_parser().parse_args(sys.argv[1:]))

  File "/home/bankhead/topBfx/bankhead/software/hotnet2/20180626//HotNet2.py", line 85, in run

    single_runs, consensus, linkers, auto_deltas, consensus_stats = consensus_with_stats(args, networks, heats)

  File "/nfs/turbo/topBfx/bankhead/software/hotnet2/20180626/hotnet2/consensus.py", line 38, in consensus_with_stats

    runs = [ (n, h, permuted_single_runs[(n, h)][i]) for n, h in network_heat_pairs ]

IndexError: list index out of range

  My call to hotnet2 is: 

python /home/bankhead/topBfx/bankhead/software/hotnet2/20180626//HotNet2.py -nf /home/bankhead/topBfx/bankhead/software/hotnet2/20180626//paper/data/networks/hint+hi2012/hint+hi2012_ppr_0.4.h5 -pnp /home/bankhead/topBfx/bankhead/software/hotnet2/20180626//paper/data/networks/hint+hi2012/permuted/hint+hi2012_ppr_0.4_##NUM##.h5 -hf intermediate/mut.gene2freq.json intermediate/cn.gene2freq.json -o intermediate/out5 -cp 3 -np 10 -hp 10 -c 16 --verbose 4

  

  I have reduced permutations to a minimal # to facilitate troubleshooting.

  Do you have any suggestions?

    Armand

[matthe...@brown.edu]

Hi Armand,

Apologies -- this is a bug that arises when the numbers of heat permutations and consensus permutations are different.

Please either pull the latest version of the HotNet2 from the HotNet2 repository, which we have just updated to fix this bug, or change line 37 of the hotnet2/consensus.py script from

for i in range(args.heat_permutations):

to

for i in range(np):

to fix the bug directly.

Best,

Matt