JMP Genomics - P-value criteria for Basic RNA-Seq Workflow

[#1]

When using the Basic RNA-Seq Workflow in JMP Genomics, how is significance defined?

Although I have specified the FDR p-value adjustment, the volcano plots still graph the unadjusted p values on the y-axis. Speaking of volcano plots, how is the horizontal line determined? This value seems to fluctuate between experiments, but I cannot find any indication of what it represents.

Under Output Data, it is not clear if the ANOVA results are filtered on the raw p-value or the adjusted p-values.

[#2 · Accepted solution]

Hi @LouisAltamura ,

The un-adjusted -log10(p-values) are plotted on the y-axis of the volcano plot. The horizontal line drawn is the adjusted (in this case FDR adjusted) p-value cutoff defined in the Basic RNA-Seq Workflow dialog (see Multiple Testing section in the figure below). This cutoff is what is used to filter the transcripts so any transcript that has a raw -log10(p-value) greater or equal to this FDR adjusted cutoff is kept and the rest are filtered out.

Here is the online help regarding some more details the report output of the ANOVA results:

https://www.jmp.com/support/downloads/JMPG101_documentation/Content/JMPGUserGuide/OT_G_EX_0021.htm#e...

Hope this helps,

[#3]

Thanks Chris! I am still using version 10.0 and the help dialog points to an older resource. This new user guide for 10.1 seems to be a lot clearer.

[#4]

Thanks @LouisAltamura ,

There is an update to 10.0. It is a hotfix that will bring your version to 10.1. You can check it out here:

https://www.jmp.com/support/notes/67/654.html 

Glad I could help,