We really want "padj" or "fdr", not just p-value

[mccalluc]

No description provided.

[mccalluc]

@gmnelson : I'm guessing that more than one of these could be present, and that the preferences are ordered? And for p-value, we're taking the negative log... is that behavior also desired for these other options?

[mccalluc]

geoff notes:

so, I've seen tools produce output with "padj" and "q_value" -- both of these should mean the same thing

have you seen "fdr" in an output, perhaps from Meeta?

I wouldn't think you would see more than one of these fields in a given output
and yes, we'll want -log10 of whatever adjusted p-value is provided in terms of making the volcano plot

I myself have not seen "fdr" but I think she told me it was a possibility. I didn't know about "q_value"... reopening to handle that case.

[gmnelson]

@mccalluc Sounds good. Even if "fdr" is used, I still don't think we'd see more than 1 of these terms used in a given output. So I wouldn't worry about establishing priority amongst the terms until we observe an output that would require that.