Hello,

You offer a conversion of the p-values into t-values.
If no p-value, nor z-value, nor t-value is provided, can we use the threshold p-value of the whole analysis (e.g., p < 0.00001 uncorrected) ?
And if so, I guess we cannot use the corrected threshold p-value. Or maybe is there a way to convert corrected to uncorrected p-value?

Thanks for your help.

Using the threshold p-value is a safe way to proceed, though conservative (leading to false negative results). Another option is to just type "p" (without quotes) for positive peaks and "n" for negative peaks, and to add a column specifying for each study whether it used a corrected or an uncorrected threshold.

I guess that conversion of corrected to uncorrected p-values is challenging, requiring unknown parameters when possible.