{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/positive-unlabeled-convolutional-neural","title":"Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs","arxiv_id":"1803.08207","date":"2018-03-22","proceeding":null,"authors":["Tristan Bepler","Andrew Morin","Julia Brasch","Lawrence Shapiro","Alex J. Noble","Bonnie Berger"],"abstract":"Cryo-electron microscopy (cryoEM) is an increasingly popular method for\nprotein structure determination. However, identifying a sufficient number of\nparticles for analysis (often >100,000) can take months of manual effort.\nCurrent computational approaches are limited by high false positive rates and\nrequire significant ad-hoc post-processing, especially for unusually shaped\nparticles. To address this shortcoming, we develop Topaz, an efficient and\naccurate particle picking pipeline using neural networks trained with few\nlabeled particles by newly leveraging the remaining unlabeled particles through\nthe framework of positive-unlabeled (PU) learning. Remarkably, despite using\nminimal labeled particles, Topaz allows us to improve reconstruction resolution\nby up to 0.15 {\\AA} over published particles on three public cryoEM datasets\nwithout any post-processing. Furthermore, we show that our novel\ngeneralized-expectation criteria approach to PU learning outperforms existing\ngeneral PU learning approaches when applied to particle detection, especially\nfor challenging datasets of non-globular proteins. We expect Topaz to be an\nessential component of cryoEM analysis.","url_abs":"http://arxiv.org/abs/1803.08207v2","url_pdf":"http://arxiv.org/pdf/1803.08207v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"positive-unlabeled-convolutional-neural","repo_url":"https://github.com/tbepler/topaz","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.08207","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}