Hi, I use Modeller 9v5 for modeling protein-DNA complex. I include DNA during modeling. I wanted to try with optimization of protein-DNA interface using a.md_level = refine.slow_large and noticed that although the protein part changes significantly comparing to the template, the DNA part stays very similar to the template. Should I use another method for turning on the optimization of DNA itself? Or better I should not try with optimization of DNA using Modeller ;-) ? Cheers, Janek Kosinski