protein of my interest have 390 amino acid. It shows only 22 % similarity to nearest pdb structure. but ab initio allows modelling of 200 residues. Suggest me a solution to model this protein