Hello again.
I am now making new models with only symmetry restraints on the alpha carbons. However, I still think it's strange that when i have a model with 26 or even 38 bad clashes they are all 2.5 or 2.6 Å. Do you know why this is? Another thing is that the alignment should not be poor. There are only three small deletions and one large insert and they are all in surface exposed loop regions. My alignment has a 57 % identity so it could not be drastically changed. I also do not think that the bad contacts are in positions close to the deletions/inserts and moving these would greatly decrease the % identity. I guess I could use just my automodel models since they are rather good but the template is fully symmetric and it's for my Master's thesis so I want to do as nice models as possible.
Thank you for putting such effort in helping all of us
/Niklas