Reduce stringency of driver prediction #1
Description
Hi, I am trying to run OncoImpact on my own data using my own gene interaction network which is smaller than the default one used by the algorithm. I have gotten the program to run as intended, but my output driver gene lists are very small (0-3 drivers predicted per sample). I suspect this is due to the smaller network I am using and likely related to the values for hub size thresholds and path lengths. Could you please advise where in the code I can find the variables to adjust these thresholds, or alternatively, how you would go about reducing the stringency of the predictions? Thanks for your help.