I have been trying to get 2D classes for my data set using cisTEM but every time I run the classification job I only get gray boxes (see attached image). I used a small subset of my data (~30 images) and was able to get reasonable class averages so I know the parameters I am using work. My collaborator has also worked with the same data set in cisTEM and was able to get good 2D averages so I don’t suspect problems with my data set. I have had several instances where cisTEM has closed unexpectedly due to power problems with our workstation and/or the VNC viewer. When opening cisTEM and returning to a project I receive the following message:
“Each database should only be opened by one instance of cisTEM at a time, otherwise corruption is possible. If it exists, you should close the other instance before continuing, if this message is the result of a crash etc. then you can override and continue.
Do you want to overide?”
I selected “override” so that I could open the project to continue working but now I am worried this has caused a file corruption that is causing the problems with the 2D classification. Is that possible? Any insight into why I am not getting reasonable 2D averages and getting these gray images instead is much appreciated. Thank you.
The reopening of the database should be fine, it should not cause this. How many rounds of 2D classification was this after? Do they always look like this, even in the initial round? Was the refinement package created by cisTEM, or did you import it?
I likely took this the screenshot after 20 rounds of classification. It looks the same after the 2nd round and the 20th round. The initial round of classification looks slightly different with a darker gray background and a whitish/ lighter gray circular shape in the center. Each of the boxes for the different classes look the same though and not what it looked like for the initial round of classifcation when I was able to get reasonable averages from ~30 micrographs. I have only used refinement packages created by cisTEM for the 2D classification. Thanks.