Single particle analysis is generally used to determine the 3D structure of globular proteins and protein complexes at high resolution.
Grids holding purified protein samples are plunge-frozen in liquid ethane and imaged. Single protein particles are oriented in various directions, allowing the 3D structure of the complex to be reconstructed by image processing methods. Filaments require a different approach for processing images; e.g. helical reconstruction.
The protein particles are imaged in a close-to-native environment. Only a small amount of protein solution is required. Crystalization is not necessary. The entire 3-D space is sampled. High resolution (~3Å or better) is reachable for 'large' proteins (mass >100kDa).
High resolution is not as easily reachable for small proteins (mass <100kDa) although this is improving rapidly. The resolution might also be limited for proteins with a high degree of variability and flexibility such as protein complexes.
Pure, stable and homogeneous protein samples; concentration higher than 0.2 mg/ml is generally required. At this concentration, we need minimally ~ 50 microliters of the solution. We would like to see a protein gel, and require full information about the buffer as well as anything known about protein stability. Buffers should ideally be low in salt <150 mM, not contain any cryo-protectants or sugars such as glycerol or sucrose. Additives such as DMSO and detergents should be avoided if possible.