search

Nesvilab / FragPipe-Analyst

An easy-to-use and interactive web application for FragPipe
https://fragpipe-analyst-doc.nesvilab.org/
0 stars 1 forks source link

Differential expression (limma) doesn't seem to use the median normalized data #41

Closed RC2754 closed 2 weeks ago

RC2754 commented 1 month ago

The normalization option, whether it's "No normalization" or "Median centered", the DE_results.csv output is exactly the same...Why isn't median centered data being used as the input for limma.

Edit: the variance stabilizing normalization option too. It changes the density plot for the QC plots but it doesn't change anything about how the DE_results.csv is calculated.

hsiaoyi0504 commented 1 month ago

@RC2754 Can you check which version of FragPipe-Analyst you are using? Also, what kind of data you are using? There is some versions ago we fixed #22.

RC2754 commented 1 month ago

I was using (http://fragpipe-analyst.nesvilab.org/) because that's what shows up on Google, and what Fragpipe v20.0 links to.

But I just tried (https://fragpipe-analyst.org/) and same issue. No normalization or VSN gives exactly the same DE_results.csv

Label-free DDA data.

(https://fragpipe-analyst.org/) takes ages to load btw.

hsiaoyi0504 commented 1 month ago

Do you use the imputation?

hsiaoyi0504 commented 1 month ago

We recently fixed one here: https://github.com/MonashProteomics/FragPipe-Analyst/commit/04a456a566057ed60abf4b4773cc2baccf9e6df9. It should be in the v0.43 that we plan to release soon.

RC2754 commented 1 month ago

Thanks. No I don't use imputation.