Hello CryoSparc team
I have been struggling with one of my datasets. I am trying to locate a small membrane protein in a bigger complex to know more about its function. I have labelled that small membrane protein subunit within the complex with a monoclonal Antibody against it.
I see an extra density for the antibody both at 2D classfication and 3D classification stage (Heterogenous refinement). But the density for the antibody becomes extremely weak and noisy once I do the consensus refinement.
Also, the noisy density for the antibody shows up even after the detergent signal after consensus refinement. I feel there is compositional heterogeniety as well as flexibility leading to this problem.
Can you guide me how to tackle this issue? I would appreciate any advice.
Thanks a bunch!
1 Like
Take all heterogeneous refinement particles from classes which have the antibody, then run the “3D classification” job (not het refine). Do not provide a volume. Set classes to 10-20, filter resolution to 14, class similarity to zero, and solvent mask near to 30 solvent mask far to 50. this should do a decent job at sorting compositional and conformational heterogeneity, especially if you have a very large datset. Generally it’s better to use a NU-refine as input but you shouldn’t have a problem skipping that step. Then pick the best class(es) and NU-refine them alone. Build a mask centered around the antibody and antigen, run local refine etc. etc. can pool similars, rerun, run 3D classification again but with the mask supplied, run 3DVA with 3DVA Display in “cluster” mode etc. etc.
2 Likes