hi! can I apply IQR(Interquartile range ) on the p-values instead of the Expression values? for genes that have several p-values by several probe. in fact, I calculated probe-IDs p-value (by limma package) in a gse that each gene has several probe.
Whether the results will be wrong?
1 answer
can I apply IQR(Interquartile range ) on the p-values
No, I don't think that summary statistics like mean and IQR make sense applied to p-values.
for genes that have several p-values by several probe.
You might be interested in this post as well How to define the gene value if a gene is represented by multiple probe sets
By the way, there are methods to combine pvalues like Fisher's method (not sure though whether different probes from the same gene can be considered independent).
Log in to answer this question.