How do I select the best pose when the L-RMSD calculation is not posibilbe as refrence structure is not available ?
In principle based on the HADDOCK score.
Check also our antibody-antigen tutorial at Education & Tutorials of the Bonvin Lab – Bonvin Lab
1 Like
what is the best/suitable HADDOCK score?
How can I select the best model from top 4 models within a cluster? What are their individual scores within a cluster?
The HADDOCK score itself! 
The first model of a cluster is always the best one in score.
But you might look at all of them to get an idea of the variability within one cluster