Perform DEGs with pyDESeq2 in CCLE Melanoma Data But Contains Negative Values

Question

[mete.han.celebi]

Hello everyone,

I would like to consult one thing about performing a DEGs analysis on CCLE melanoma data. I have a pandas df where the rows are cell lines, and columns are the gene names. In Figure-1 you can see a screenshot of that df which contains negative values for some genes in this CCLE melanoma data. I think it arises since CCLE data is a log2 transformed.If I am wrong, please correct me :)

I created a metadata where I define a condition for each cell line. Here is my metadata:

I setten up the pyDESeq2 as follow:

However I got an error, most importantly about the negative values and non-integer values. So my question is that in order to be able to perform DEGs analysis between the conditions (lower_03 vs higher_1) that I defined in the metadata, what sort of strategy would you recommend? Should I take undo log2 of the df and make the values as integer? What would you suggest as a best strategy?

Thank you very much in advance, and looking forward to learning from you!

Answer 1

[ATpoint]

negative values for some genes in this CCLE melanoma data. I think it arises since CCLE data is a log2 transformed

Nobody can tell you. You need to consult the documentation from where the data come from. DESeq2 expects raw integer counts to obey its model assumptions.

[mete.han.celebi]

Thank you so much for your reply! I will check it now!