Visualize one-tailed test data

Question

[szabo.marton]

Hi all,

I wondering whether it is possible to plot p-values from of one-tailed test against fold change values, eg.: something similar like Volcano plot? We performed some proteomics experiments and TOST (two one-tailed t test) were performed for statistical analysis. Since we've got p-values greater than 1, general visualization methods are not suitable for such data.

Any suggestions are appreciated.

[seidel]

Regardless of having p-values > 1, is the interpretation the same? i.e. smaller = more desirable than larger? If so, you could still make a Volcano plot using the -10log(p) trick y axis. That is, if you want to show the relationship between degree of difference and a significance score. (If I understand your question correctly).

[szabo.marton]

Thank you for your response. Yes, the lower the p-value, the greater the observed difference (roughly..). If I have p-value greater than 1, the -log10 transformation will result negative values, which make the Volcano kinda strange. That is why I asked this question.