DOE prototype OS 1.6 vs EP 23-1 discrepancies

edit

Hi, Beginner here and I am running DOE models (medium office in Buffalo) in EP and in OS for validation purposes and getting different End Use values for cooling heating etc.. Errors can be up to 10%. When I run the models as is, errors are only 1-3%. But I am changing U values and SHGC of windows in both software and comparing that is when the error gets higher.

In OpenStudio I am using the Prototypes from the BCL measure. While in Energy Plus I am using the corresponding IDF files from energy.gov. Is that what is creating the discrepancies? and what would be the best way to reduce those errors?

Thanks!