eman2:e2gmmoriginal
Differences
This shows you the differences between two versions of the page.
| eman2:e2gmmoriginal [2025/06/19 01:27] – created steveludtke | eman2:e2gmmoriginal [2025/06/19 01:27] (current) – steveludtke | ||
|---|---|---|---|
| Line 32: | Line 32: | ||
| </ | </ | ||
| - | Now we begin training the Gaussian model from scratch. We start from a tiny box size and a small amount of Gaussian functions so it converges faster. Then run with a Fourier box size of 48 so the Gaussian model matches the density map at ~15 �. When increasing the number of Gaussian functions, the new Gaussian will be seeded near the old ones, so the overall shape of the Gaussian model is preserved at each run of the refinement. The mask will exclude the Gaussian centers outside the volume. The last command can be repeated if the Gaussian model does not match well with the density map. | + | Now we begin training the Gaussian model from scratch. We start from a tiny box size and a small amount of Gaussian functions so it converges faster. Then run with a Fourier box size of 48 so the Gaussian model matches the density map at ~15 Å. When increasing the number of Gaussian functions, the new Gaussian will be seeded near the old ones, so the overall shape of the Gaussian model is preserved at each run of the refinement. The mask will exclude the Gaussian centers outside the volume. The last command can be repeated if the Gaussian model does not match well with the density map. |
| < | < | ||
eman2/e2gmmoriginal.1750296421.txt.gz · Last modified: by steveludtke