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

Fragpipe analyst normalisation has no effect for DIA data #22

Closed TBDerm closed 3 months ago

TBDerm commented 6 months ago

Hi all,

I'm running a proteomics experiment and noticed something about the 'no normalisation' / 'variance stabilizing normalisation' option in fragpipe analyst.

I'm using DIA intensity data (output from DIA-NN). I have two datasets: the whole set with all detected proteins, and a subset of ~45 proteins. There are 3 replicates of 2 conditions.

So the mystery lies here: I have tried uploading both the whole dataset and the sub-dataset, with and without the normalisation option. Regardless of dataset, normalisation has no effect on the outputted fold change and adjusted p-values. HOWEVER, a colleague has a similar situation: whole dataset (DIA-NN output) and a subset of certain proteins from this - when they try these different settings on Fragpipe analyst, normalisation makes no difference to the sub-dataset output, but when the whole dataset is inputted, selecting normalisation does produce different fold change values for some proteins, vs when no normalisation is selected.

So this has left us all confused, and our supervisors are suggesting we manually normalise the DIA-NN output before inputting it to fragpipe analyst. Does anyone else have suggestions on getting around this issue, or if it even is an issue? With the differences occurring only sometimes, I do not know if it is working as it should, or if there is a problem.

Does anyone know why fragpipe analyst is doing this, or how the built-in normalisation works? Any insights would be appreciated!

hsiaoyi0504 commented 5 months ago

Thanks for the feedback. @TBDerm If you have the data and able to share with me yihsiao@umich.edu, I can take a look or I currently don't have much idea about it.

hsiaoyi0504 commented 4 months ago

@TBDerm I think it fixed now. I realized that vsn was not applied for DIA previously. Would you mind checking it again?

hsiaoyi0504 commented 3 months ago

Close given no response.

veslugleynat commented 3 months ago

Hi, Thanks for fixing vsn for DIA data. Is it possible to manually replicate what it does to check my data please? Thanks!

hsiaoyi0504 commented 3 months ago

@veslugleynat Can you elaborate a bit what's your request? I am confused.