Welcome to the MapMan Family of Software Forum

Please do not hesitate to register and post your question.

Don't forget to subscribe to your posted message so you get notified on updates.
Every question you post will help others and or enhance the software!

Post a question,   post a bug!

Welcome to the MapMen Family of Software Forum Welcome to the MapMen Family of Software Forum

MapMan Features

about P-value

about P-value
Answer
6/24/09 11:38 AM
Hi,
 I have a few questions about the Wilcoxon rank sum test in this software.
1. How does Wicoxon test identify the major bins of genes? Does it use Log fold-change of my uploading data?
2. If the p-value of the Wilcoxon rank sum test is more than 0.05, that will mean the BIN function would not be true. Is that right?

Re: about P-value
Answer
6/26/09 10:04 AM as a reply to Dear yw.
The Wilcoxon test:

You can imagine it to be conceptually similar to a t-test, however it uses ranks instead of the underlying data and thus doesn't need (approximate) normality of data.

It then tests if the changes within a BIN are more extreme than the changes in all the remaining genes. However, as said before this is done on a rank scale.


An example:
BIN 1, 3 genes with log2 FC
gene 1 -2
gene2 -3
gene3 -0.5

all other bins 4 genes
geneA 1
geneB 0.2
geneC -0.2
geneD 2

ranks BIN1
gene1 2
gene2 1
gene3 3

ranks rest
geneA 6
geneB 5
geneC 4
geneD 7

A t-test would evaluate this as barely significant p=0.046, whereas the Wilcox test would not detect any significant change (p=0.057).

Please note that like always the failure to detect a significant change doesn't necessarily imply that there is no change.

Best
Björn