Equivalent of Fold-Change analysis?

[joannawolthuis]

Hi there!

Thanks for your hard work on this package!
I'm trying it out on my metabolomics data currently with a high amount of missing values.
I was wondering if there was a way to make an equivalent of a volcano plot using your Wald test alongside another (not currently implemented?) test that tests something similar to fold change between conditions :)

Kind regards and thank you for your time,
Joanna

[const-ae]

Hi Joanna,

just to make sure that I understand what you are asking: do you want to perform some alternative test and use the resulting p-value on the y-axis of a volcano plot?

[joannawolthuis]

Hi, thank you for your quick reply! You're correct , oftentimes people combine a t-test with a fold-change test looking at the magnitude and direction of effect... However natively these have trouble with missing values.

[joannawolthuis]

To add: I'm asking here rather than using the completed table because I saw your comment elsewhere on that that might show an unrealistic amount of significant features :)

[fedeserral]

hi, i have the same question.

Thaks!

[const-ae]

Sorry, I am still a bit confused about this question. Is this about performing a statistical test where you check that the log-fold change is larger than some value? Similar to the lfcThreshold argument in DESeq2 (https://www.bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#tests-of-log2-fold-change-above-or-below-a-threshold)?

[fedeserral]

yes, but is any way to do it using proDA? I didn't find the way.

[const-ae]

No, I haven't implemented such an argument. One would need to modify the run_wald_parameter_test to do something like the DESeq2::results , but I currently don't have time to implement such a feature. But I if you are interested in making a PR, I am happy to help :)

[fedeserral]

thaks for the answer! It would be great this implementation to work with proteomic data.