Generation of improved 2D class averages from large single-particle cryo-EM datasets
Production of a reliable 3D starting model in a rapid and unbiased fashion
Overcoming inherent limitations in widely used clustering approaches
The method is many times faster than other widely used approaches