My protein has preferred orientation. It is easy to pick top views but side views are few and hard to pick. The only way I was able to pick more side views was by training TOPAZ (wrapper) separately for side views. Now I have 2 set of particles, top views (TOPAZ) and side views (TOPAZ).
Is there a way for me to sort out duplicate particles among these two particle sets so I don’t accidentally use the same particle twice in refinement?
I tried training Topaz with top and side views, but the majority of the particles were still top view (though there were more side views compared to other picking methods). When I trained Topaz with just side view particles, the model provided considerable increase in the side view population. Therefore I decided to perform separate Topaz extraction using two different models.
Topaz has been excellent. I was about to give up and go back to wet lab to make more grids in order to get more side views but Topaz was able to extract 3x to 5x more side view particles compared to other picking methods I tried.
I’m glad you found a way to get those side views! Thanks for the information on a top+side view model versus top-view model + side-view model. I’ll keep this example in mind when teaching people how to use Topaz.