Particle detection is the first thing to do in the single particle reconstruction as
soon as the CryoEM temp/zeyun are digitized and/or CTFcorrected. The goal of this
step is to locate all particles from the CryoEM temp/zeyun. Since achieving highresolution
reconstruction often requires over hundreds of thousands of particles, it is extremely
important to design a fast and automatic algorithm for particle detection. We have developed
two methods for this task. One is based on data clustering and the other is based on Voronoi
diagram and distance transform.
Clusteringbased Method for Particle Picking
Details will be available later or upon request. The following pictures show two examples
of particle picking on P97 particles using this method.
Fig. 1 Two examples of particle picking on P97 particles using our clusteringbased method.
Computational Geometric Method for Particle Picking
Details will be available later or upon request. The following pictures show two examples of
particle picking on Keyhole Limpet Hemocyanin (KLH) particles using this method.
Fig. 2 Two examples of particle picking on P97 particles using our clusteringbased method.
