Hi,

I had a question about which values to use for the scoreCol when generating a dba object in DiffBind. My peak files are SICER islandSummary files, which contain columns for pvalue, fold change, and FDR. Currently, I have the scoreCol in my sample sheet set as "sicer" which is pulling the fold change values from my peak files (column 8). Is it best to use fold changes as input into DiffBind, or would pvalues or FDR be a better score to use?

Thanks for your help,

Emily

In most DiffBind analyses, the initial peaks scores aren't used for anything important. These scores are only used for making plots prior to calling dba.count() (for example, clustering or heatmaps). Once you have called dba.count(), the scores are all based on actual counts and the initial scores are ignored.