[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-664720-105":3,"detail-sidebar-cat-0-en-105":80,"doc-detail-664720-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":73,"head_meta":75,"extra_data":77,"updated_unix":79},105,"en","picking-particles-in-cryo-em-images-without-knowing-particle-size","Picking Particles in Cryo-EM Images without Knowing Particle Size","","Focuses on particle picking in cryogenic electron microscopy (cryo-EM) images, aiming to reconstruct biomolecular structures from 2D projections without requiring prior knowledge of particle size. Explains cryo-EM imaging and computational reconstruction steps, then analyzes common challenges such as low signal-to-noise ratio and confounding factors like ice contamination, background noise, and particle overlap. Reviews limitations of template-based and size-dependent DNN approaches and introduces a RetinaNet-style object detection model with focal loss plus data augmentation that varies particle sizes to improve cross-dataset generalization.",{"@graph":14,"@context":72},[15,34,55],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/picking-particles-in-cryo-em-images-without-knowing-particle-size/664720/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/picking-particles-in-cryo-em-images-without-knowing-particle-size/664720.png","ImageObject",300,407,{"name":42,"@type":43},"Helloiii","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-30","2026-10-29",true,{"@type":52,"interactionType":53,"userInteractionCount":30},"InteractionCounter",{"@type":54},"ViewAction",{"@type":56,"mainEntity":57},"FAQPage",[58,64,68],{"name":59,"@type":60,"acceptedAnswer":61},"What is the core goal of the particle picking approach described here?","Question",{"text":62,"@type":63},"To pick particles in cryo-EM micrographs and enable 3D reconstruction of biomolecular structures without knowing the particle size in advance.","Answer",{"name":65,"@type":60,"acceptedAnswer":66},"Why is particle picking difficult in cryo-EM images?",{"text":67,"@type":63},"Micrographs often have very low signal-to-noise ratio, along with interferences such as ice contamination, background noise, amorphous carbon, and particle overlap.",{"name":69,"@type":60,"acceptedAnswer":70},"How does the method avoid relying on a fixed particle size?",{"text":71,"@type":63},"It addresses cross-domain limitations by training with data augmentation that includes diverse particle sizes and patch cropping, so the detector can generalize beyond the ground-truth size used in training.","https://schema.org",{"og:url":32,"og:type":74,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":76,"canonical":32},"index,follow",{"doc_id":78,"site_id":7},664720,1793268707,{"code":4,"msg":81,"data":82},"success",[83,87,91,95,100,105,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":84,"show_sort_weight":85,"slug":86},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":88,"show_sort_weight":89,"slug":90},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":92,"show_sort_weight":93,"slug":94},"Exam",70,"exam",{"id":96,"doc_module":4,"doc_module_name":25,"category_name":97,"show_sort_weight":98,"slug":99},5,"Comic",60,"comic",{"id":101,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},6,"Technology",50,"technology",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},7,"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":96,"slug":129},19,"General","general",{"code":4,"msg":81,"data":131},{"doc_id":78,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":30,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":79,"read_time":144},5909897196768,"https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0","Picking Particles in Cryo-EM Images without Knowing Particle Size  \nYuewei Lin 1 , Xiaoning Li2 , Qun Liu 1 , Shinjae Yoo 1  \n1. Brookhaven National Laboratory  \n2. Stony Brook University  \nCryogenic Electron Microscopy (cryo-EM)  \nIn 2017, the Nobel Prize in Chemistry--“for developing cryo-electron microscopy for the high-resolution structure determination of biomolecules in solution.”  \nHow cryo-EM works  \nEach 2D image is a projection of its the 3D shape  \n– Each pixel value in the 2D image is the sum of the values along the line (along the direction of the electron beam) through the 3D sample  \nImage from [http://biomachina.org/courses/structures/091.pdf](http://biomachina.org/courses/structures/091.pdf)  \nComputational steps in cryo-EM  \nAims to reconstruct the structure of a “particle”: single molecule (e.g. , protein) or composed of many molecules (e.g. , a ribosome)  \n• Spreading identical particles, and image them using an EM. Particles are positioned with different orientations  \n• Picking as many as particles in micrographs  \n• 3D structure reconstruction using 2D particles  \nChallenges  \n• The micrographs usually have very low signal-to-noise ratio (SNR)  \nChallenges  \n• The micrographs usually have very low signal-to-noise ratio (SNR)  \n• Other interferences, such as ice contamination, background noise, amorphous carbon and particle overlap.  \n• High-resolution reconstruction requires extensive particles identification (> 100,000) .  \nLimitation of the existing methods  \n• Traditional template based models  \n• Very slow (minutes per micrograph)  \n• Need to know particle size  \n• Sensitive to noise  \n• Deep neural network (DNN) based models  \n• Slow (seconds to tens of seconds per micrograph)  \n• Need to know particle size  \nAdvanced DNN based object detection model--RetinaNet  \nStructure:  \n• Resnet for encoder  \n• Feature pyramid network for multi-scale feature extraction  \n• Subnets for class and box regression in each scale  \nFocal loss:  \n• Penalizes less to easy samples  \n• Deals with class imbalance  \nFail to apply across domains  \n• All the ground truth particles are the same size  \n• Model tends to pick particles with the same size as ground truth  \n• It impedes the trained model to apply on different datasets  \nTrained on particle size 200 Tested on particle size 120  \nData augmentation with diverse size of particles  \nUtilizing different sizes of particles as training data  \n• Cropping patches with random positions and sizes  \n• Re-sizing them to be the same size (e.g. , 1000×1000) .  \n4096×4096 1388×1388 1905×1905 2643×2643  \nThe key to make the model picking particles without knowing the size!","cbCaifSrnpBUF1oM","https://ap.wps.com/l/cbCaifSrnpBUF1oM","pdf",3719582,21,"English","# How cryo-EM works\n## Computational steps in cryo-EM\n# Challenges and limitations\n## Limitations of existing methods\n# Advanced model and training strategy\n## RetinaNet structure and focal loss\n## Cross-domain generalization without fixed particle size\n# Data augmentation strategy","[{\"question\":\"What is the core goal of the particle picking approach described here?\",\"answer\":\"To pick particles in cryo-EM micrographs and enable 3D reconstruction of biomolecular structures without knowing the particle size in advance.\"},{\"question\":\"Why is particle picking difficult in cryo-EM images?\",\"answer\":\"Micrographs often have very low signal-to-noise ratio, along with interferences such as ice contamination, background noise, amorphous carbon, and particle overlap.\"},{\"question\":\"How does the method avoid relying on a fixed particle size?\",\"answer\":\"It addresses cross-domain limitations by training with data augmentation that includes diverse particle sizes and patch cropping, so the detector can generalize beyond the ground-truth size used in training.\"}]","Picking Particles in Cryo-EM Images without Knowing Particle Size | PDF",53]