Scoring/ranking problem in basicmhc1 and tepitope

Hi! I’m using tepitope and basicmhc1 models and i noticed the cutoff method in tepitope work well but not in basicmhc1. In particular, with basicmhc1 epitopes with rank 1 are the worriest ones and have low score. Better epitopes are those with higher ranks. When using cutoff=0,95, best epitopes are removed. How this issue can be fix? Thank you 
Francesco 