Hi all, I would appreciate some thoughts from those with experience in protein structure prediction. I have already had some excellent advice on how to predict free model (ab initio) structures given the amino acid sequence and almost no template/homology modelling using free services such as I-TASSER. Now I would like to know what can be done with the structures that have a 'good' c-score e.g. close to 0 (where c-score is a range [-5,2])?

I have taken the 'good' structures, solvated them, added ions if needed and performed a thorough molecular dynamics simulation under physiological conditions (although, water density is an unknown so I used a fully solvated model).

We essentially have 105 sequenced genes. We've never seen them before until now and they do not exist in any database. Most homology e.g., via PSI-BLAST, returns very bad results. We simply want a good guess as to what they could be doing from our 'good' predicted models. 

Any advice would be appreciated. Thanks.