I’m working on a doughnut shaped molecule. Using Topaz I was able to pick >400,000 particles from a combination of 0-degree tilt and 15 degree tilt images, the latter collected to try to get more side views. The particles refined to 3.2 A resolution model but the preferred orientation problem was only somewhat lessened. I then created templates to focus on side views and picked a set of particles that initial model and 2D classification reduced to about 200,000 particles that refined to 3.3 A structure. The distribution is improved for that set. Here is the problem - when I used ‘remove duplicate particles’ to combine the two sets of particles from Topaz and Template picking, I got a new particle set (which removed about 80,000 of the particles) that fails when I do 2D classification or initial model jobs. The error is ‘all particles must have the same data sign’. Does a different sign get applied when using Topaz vs Template picking from the same exposures? Any way to fix this problem by changing the sign for one of the two sets prior to combining them via remove duplicates job?
Data sign problem upon using removing duplicates to combine Topaz and Template picked particles
Welcome to the discussion forums! Did you happen to run Local Motion Correction on one of the subsets? This job has an older convention that outputs particles with the opposite
blob/sign to that of the patch motion job. One way to fix this is by running a Downsample Particles job with the option ‘Flip data sign’ turned on, and the crop size set to the same value as your extraction box size (i.e., no cropping/downsampling).
Thank you Valentin. I did to Local Motion Correction. What you describe sounds like it should work. I’ll be giving it a try!
That did the trick! Thanks a million!