Particle Detection in Cryo-Electron Micrographs

A new learning-based approach is presented for particle detection in cryo-electron micrographs  using the Adaboost learning algorithm. The approach builds directly on the successful detectors developed for the domain of face detection. It is a discriminative algorithm which learns important features of the particle’s appearance using a set of training examples of the particles and a set of images that do not contain particles. The algorithm is fast (10 seconds on a 1.3 GHz Pentium M processor), is generic, and is not limited to any particular shape or size of the particle to be detected. The method has been evaluated on a publicly available dataset of 82 cryo-EM images of keyhole lympet hemocyanin (KLH). From 998 automatically extracted particle images, the 3-D structure of KLH has been reconstructed at a resolution of 23.2 °A which is the same resolution as obtained using particles manually selected by a trained user.

Results

I

Left and Right : Qualitative results on particle detection  on the top and sides views of KLH is shown. The white dots represent the locations of detected particles. 

Center: The curve shows the receiver operating characteristic (ROC) of the final particle detector. Our1 is the ROC with circular particles removed and Our2 is the ROC obtained without removing the circular particles before detection of the rectangular particles.

Publications: 

• Satya P. Mallick,  Yuanxin Xu and David J. Kriegman  "Detecting particles in cryo-EM micrographs using learned features," Journal of Structural Biology, Jan/Feb 2004, Volume 145, Issues 1-2.