I am new user in CryoSPARC. I am working on a reconstruction of a SPA data set recently collected on a Talos Artica 200Kve using Falcon III (K3 is temporarily out of service) in linear mode. It was very first time I am doing SPA on the sample. My sample should look like a ring structure according to negative staining (attached). My sample was frozen on QF Cu, 300mesh 1.2/1.3 with 2nm continuous carbon film and it has preferred orientation (mostly top or bottom view). But it seems that I have ~1% tilted views, Ab-initio result (screen shot) looks a bit strange, anyone has some suggestions to adjust the parameters?
My goal for this dataset is not to get high resolution and it was a good sample for me to get practice on doing SPA data collection and process.
Given that you know from 2D and NS results that you likely have 12-fold symmetry, have you tried imposing C12 symmetry during ab initio? This is not generally recommended, but it can give good results in certain cases, and in this case it seems like you have a fair handle on the likely symmetry from 2D.
Also you might want to play around with the initial and final resolution values - sometimes starting at higher resolution (e.g. 12 or even 9 Å) can give good results. With small, featureless membrane proteins that have good high resolution features, often starting at 9 Å and going to 7 Å gives good results.
I jsut realized that my structure contains 11-ring (most top views) and 12-ring (slightly tilted views)… It might be another reason that Ab initio did not give me a good initial average. I am going to try multiple classes in the Ab initio reconstruction step.