To know the scores

[ABanaeiEsfahani]

Hello,

I have a question regarding the scores we get in the output file including "dens_string", "meanp_meme" and "resid". Could you please tell me where I can exactly find some description about them? Even in the original paper, I could not find so much about them. It would be great if you can a little bit tell me about and what would be a sensible cutoff to filter the bi-clusters accordingly. Thanks a lot.

Best regards,
Amir

[weiju]

Hi,
I am assuming you refer to thefile cmresults-postproc.tsv.bz2. For us this has no other use outside
of using it to get more information for possible debugging. I believe that the R version previously
generated it so it is generated for compatibility reasons and it simply is a copy of the statistics that are contained in the tables iteration_stats and cluster_stats in cmonkey_run.db

[ABanaeiEsfahani]

Hi,

Thanks. Yes I know that it's a copy of the statistics in cmonkey_run.db. What do they mean statistically? For examples p,q value in statistics are pretty obvious but I could not find so much info about resid. In the Mtb paper (Peterson E et al) They filtered the bi-clusters based on resid value (<0.55). I guess I should know what it means exactly before setting a cutoff. Does that make sense?