Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

- JMP User Community
- :
- Discussions
- :
- table of fold change P value and FDR for comparison of groups

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

table of fold change P value and FDR for comparison of groups

Sep 5, 2018 10:34 AM
(1816 views)

I have datasets with many response variables (~15K) and many samples (~7K). I have clustered the samples based on prior knowledge into 10 clusters, and need to define the variables that are characteristic for each. I want to generate a table/matrix that lists all the variables vs the clusters, with the fold change (ratio) in mean response values (each cluster vs all other samples together) with P values (either corrected or not) and FDR if possible.

Hoping there is a simple way to generate this. ANOVA or negative binomial options would be helpful.

3 REPLIES 3

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: table of fold change P value and FDR for comparison of groups

I would suggest you look at the screening tools: Analyze > Screening > Response Screening. I often use the response screening platform for what I think you are trying to do. The documentaiton is at:

https://www.jmp.com/support/help/14/response-screening.shtml

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: table of fold change P value and FDR for comparison of groups

Response screening gives the difference in means. However for our purposes the difference in mean values is not as relevant as the difference in the ratios of each gropup mean to the whole.

Hence a table of fold change with P values and FDR is needed. I wonder if their are availble scripts that do this.

thanks

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: table of fold change P value and FDR for comparison of groups

My experience in such cases was in virology when the useful information was a doubling but the response was titer. So a transformation could make the response more meaningful transforming as Ln( titer ) / Ln( 2 ). (That is effectively the logarithm base 2.) You can do this transformation with a column formula without the need for a separate script.

Learn it once, use it forever!

Article Labels

There are no labels assigned to this post.