Symmetry code doesn’t work

[rakeshr10]

Hi @ZiyaoLi ,

I tried the symmetry code example based on your colab notebook. It does n’t seem to work. If I give a homotetramer, the PDB chains seem to be far apart in space than being close together in tetrameric cyclic symmetry.

I installed libmsym package which was a dependency, though I did n’t find any installation instructions for this in the repo.

Would be great if you could look into this.

Regards
Rakesh

[ZiyaoLi]

If the model gives far apart tetramer, it indicates that the model tends to think the input sequence cannot form a tetramer.

You may check if the monomer structure is of anticipation.

libmsym is not required. It's only used in training.

[rakeshr10]

I tried your colab notebook example. When I ran without libmsym, it was complaining about this dependency.

The monomer structure based on your example sequence in the colab notebook seemed fine to me.

[ZiyaoLi]

Thanks for the reply. I'll remove the dependency.

[rakeshr10]

Hi @ZiyaoLi ,

Did you remove the dependency for libmsym. I don’t see any changes being made in the repository.

Regards
Rakesh