Microtubule Seam-search
Microtubule seam-search is very important to separate α- and β-tubulins.
What the seam-search algorithms in cylindra
do is simply adding a new column named
"isotype-id" to the molecules features. As seam-search is usually combined with
subtomogram averaging and alignment, these methods are located in the STA widget.
Seam-search Based on Cross-correlation
API: seam_search
GUI: STA widget > Analysis > Seam search > Seam search by correlation
Although the structure of α/β-tubulins are very similar, it is known that trying all the possible patterns and comparing the cross-correlation is useful to distinguish them.
List of parameters
- "layer" is the molecules layer to be seam-searched.
- Set the template image and mask parameters in the STA widget.
- "anti-template path" is the path to the anti-template image.
- "interpolation" is the interpolation method used for resampling the sub- volumes.
- "npf" is the number of protofilaments. If molecules are created in
cylindra
, this value should already be determined by CFT. - You can select how to (or not to) show all the averages for different seam locations by the "show averages as" combobox.
- "cutoff" is the relative cutoff frequency of the low-pass filter. This value is usually smaller than the value used for subtomogram alignment.
Molecules labeled with 0
are the molecules that should be considered as the template,
and 1
the anti-template.
Manual Seam-search
API: seam_search_manually
GUI: STA widget > Analysis > Seam search > Seam search manually
This method labels the molecules by the given seam location. location=0
will label
molecules with
[0, 0, ..., 0,
1, 1, ..., 1,
0, 0, ..., 0,
1, 1, ..., 1,]
and location=1
will label molecules with
[1, 0, ..., 0,
0, 1, ..., 1,
1, 0, ..., 0,
0, 1, ..., 1,]
Seam-search by Features
GUI: STA widget > Analysis > Seam search > Seam search by feature
This method labels the molecules by the given feature. The feature should be such that approximately label the α- or β-tubulins. Usually, molecules should be labeled by classification using microtubule-associated proteins as the fiducials.