Hi everyone! I am currently using limma to analyse my peptides datasets, I usally transform my data to log2 before performing limma, but I notice that if I don´t transform them, I get different and not concordant results, why is that? Which results should I consider the most "reliable" and why?

Have a nice week!

P.S. I also read in papers that transforming the data to log2 is a common practice with omics data, that is why I usually do it.