Hi DepMap team,
Seeking some clarification over the EffectSize which is referred to as “PosteriorMean” in one of the previous posts and in R package cdsrmodels. I am looking at a rather strange distribution of PosteriorMean for a specific Two Class comparison. The first plot enclosed is made using the R package cdsrmodels, but this is also reproducible on the DepMap web portal using Two Class comparison (for which I have also enclosed a second plot). The query uses two groups of cell lines which are below. Would you be able to comment on why the EffectSizes following shrinkage of fitted betas are all (a bit oddly) centred below zero, when the fit[[“res.table”]]$betahat were centered on zero?
Group 1:
ACH-000867,ACH-000785,ACH-000866,ACH-000924,ACH-000718,ACH-000698,ACH-000290,ACH-000378,ACH-000161,ACH-000395,ACH-000463,ACH-000311,ACH-000150,ACH-000431,ACH-000510,ACH-000695,ACH-000781
Group 2:
ACH-000528,ACH-000769,ACH-000800,ACH-000700,ACH-000176,ACH-000587,ACH-000585,ACH-000627,ACH-000035,ACH-000638,ACH-000894,ACH-000596,ACH-000481,ACH-000871,ACH-000858,ACH-000787,ACH-000677,ACH-000012,ACH-000945,ACH-000662,ACH-000261,ACH-000681,ACH-000901,ACH-000511,ACH-002077,ACH-000975,ACH-000890,ACH-000562,ACH-002650,ACH-000187,ACH-000639,ACH-000339,ACH-000525,ACH-000853,ACH-000851,ACH-001386,ACH-000434,ACH-000514,ACH-000414,ACH-000665,ACH-001549,ACH-000706,ACH-000251,ACH-002526,ACH-002051,ACH-000589,ACH-000264,ACH-000843,ACH-000840,ACH-000912,ACH-000667,ACH-000747,ACH-000438,ACH-000021,ACH-000980,ACH-000845,ACH-000448,ACH-000766,ACH-000343,ACH-000774,ACH-000921,ACH-000790,ACH-000767,ACH-000454,ACH-000916,ACH-000292,ACH-000837,ACH-000861,ACH-000563,ACH-000355,ACH-000757,ACH-000416,ACH-000442,ACH-000841,ACH-000860,ACH-000929,ACH-000390,ACH-000496,ACH-000878,ACH-000257,ACH-000521,ACH-000029,ACH-002531,ACH-000869,ACH-000337,ACH-000444,ACH-000886,ACH-000888,ACH-000030,ACH-000775,ACH-000559,ACH-000143,ACH-002035,ACH-000015,ACH-000791,ACH-001075,ACH-000900,ACH-000666,ACH-000872
Thank you so much in anticipation
