A Self-Supervised Workflow for Particle Picking in Cryo-EM

High-resolution single-particle cryo-EM data analysis relies on accurate particle picking. To facilitate the particle picking process, we have developed a self-supervised workflow. Our workflow includes an iterative strategy to use the 2D class average to improve training particles and a progressively improved convolutional neural network (CNN) for particle picking. To automate the selection of particles, we define a threshold (%/Res) using the ratio of percentage class distribution and resolution as a cutoff. Our workflow has been tested using six publicly available data sets with different particle sizes and shapes, and is able to automatically pick particles with minimal user input. The picked particles support high-resolution reconstructions at 3.0 [A] or better. Our workflow offers a way toward automated single-particle Cryo-EM data analysis at the stage of particle picking. The workflow may be used in conjunction with commonly used single-particle analysis packages such as Relion, cryoSPARC, cisTEM, SPHIRE, and EMAN2.