Processing filament

Hi all,

I am currently working with a AAA+ ATPase that comes off of size exclusion columns in multiple peaks, one of which contains filaments of varying length. The protein usually forms a hexameric ring, which each of the monomers containing a D1 and D2 domain connected by a linker, but its relatives have been known form helical filaments or to dodecamerize by forming a structure where two hexamers interact by their D2 domains. Without knowing whether the filaments I am seeing on my micrographs are helical, if they stack as D1:D2:D2:D1 or another pattern, or something else, what is the best way to go about processing this dataset?