Electron Microscopy 4 Flashcards Preview

M3D > Electron Microscopy 4 > Flashcards

Flashcards in Electron Microscopy 4 Deck (11):

What are the four parameters that affect resolution?

1. Electron dosage
2. Magnification and defocussing
3. Tilt angles and increment
4. Amount of data collected


What are the key issues with single particle analysis?

Low signal to noise ratio.
Ambiguous orientations of the single particles.


What does describing the CTF allow EM to do?

Reverse engineer the original object.


What is 3 methods used to pick particles manually in SPA?

1. Cross correlation
2. Neural networks
3. Edge detection


Describe cross correlation.

Choose a template from the images and use this to select other images in the same orientation.
This class of images are then averaged to produce a class average with a better signal to noise ratio.


What are the 3 Ab initio ways of generating a starting model for SPA?

1. Common Lines
2. Random conical tilt.
3. Sub-tomogram averaging


What other method can be used to generate a starting model for SPA?

Can use a pre-existing model - either similar structure or if the molecule has already been determined by another method then can use this.


Describe the method of starting from a pre-existing model.

The model must be low pass filtered (60A) - remove all the high resolution features.
Check that all the high resolution features in the new model are in many class averages (overlap), as they may have come from model.


Describe the method of generating a common lines starting model.

Take different orientations (class averages) of the structure and align them using overlap of pixels.
Then align class averages to this low pass filtered model.


Describe the method of generating a random conical tilt starting model.

The molecule is titled two different angles and images are collected. These are aligned as there exact tilt angle is known.
There is a missing cone of information (z axis) - this can be overcome by combining several tilt reconstructions.


Describe the method of generating a sub-tomogram averaging starting model.

Select many particles that have been tilted - superimpose the same tilt angles to generate an average.