[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-127712-105":3,"detail-sidebar-cat-0-en-105":85,"doc-detail-127712-en":135},{"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":78,"head_meta":80,"extra_data":82,"updated_unix":84},105,"en","a-machine-learning-model-of-perturb-seq-data-trials-final-testing-accuracy-for-each-dataset","A Machine Learning Model of Perturb-Seq Data - Trials - Final Testing Accuracy for Each Dataset","","A Machine Learning model is trained to predict gene expression perturbations on new Perturb-Seq datasets, targeting spaceflight-relevant biological systems with altered expression profiles. The workflow selects a Perturb-Seq dataset, preprocesses and normalizes data, applies Leiden clustering, and annotates clusters using GSEA to establish biological significance. A scikit-learn MLP is optimized through parameter and hyperparameter trials, then refined by prediction and quality assessment. Results show OSD-91 RNA samples matched single clusters in K562 Essential and RPE1 datasets.",{"@graph":14,"@context":77},[15,34,56],{"@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/a-machine-learning-model-of-perturb-seq-data-trials-final-testing-accuracy-for-each-dataset/127712/",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/a-machine-learning-model-of-perturb-seq-data-trials-final-testing-accuracy-for-each-dataset/127712.png","ImageObject",300,407,{"name":42,"@type":43},"Emma Wilson","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-22","2026-08-05",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",15,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69,73],{"name":60,"@type":61,"acceptedAnswer":62},"What is the goal of the machine learning model in this work?","Question",{"text":63,"@type":64},"To train a machine learning model on gene expression perturbational datasets so it can predict perturbations in new datasets, helping identify genetic sources of altered expression profiles such as those from spaceflight.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"Which steps are included in the processing workflow before model training?",{"text":68,"@type":64},"The workflow selects a Perturb-Seq dataset, performs pre-processing, filtering, and normalization, applies Leiden clustering, and uses gene set enrichment analysis (GSEA) to annotate biological significance for each cluster.",{"name":70,"@type":61,"acceptedAnswer":71},"What model and optimization approach are used?",{"text":72,"@type":64},"A scikit-learn Multi-Layer Perceptron (MLP) is selected, configured, and optimized via parameter and hyperparameter tuning trials, followed by refinement and selection of the perturbational model.",{"name":74,"@type":61,"acceptedAnswer":75},"What do the results show for the OSD-91 RNA samples?",{"text":76,"@type":64},"OSD-91 RNA samples are predicted to belong to single clusters in both the K562 Essential and RPE1 Perturb-Seq datasets, with average sample matches of 93.0% (Cluster 18) and 87.3% (Cluster 3), respectively, supported by GSEA associations.","https://schema.org",{"og:url":32,"og:type":79,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":81,"canonical":32},"index,follow",{"doc_id":83,"site_id":7},127712,1785941123,{"code":4,"msg":86,"data":87},"success",[88,92,96,100,105,110,115,119,124,127,131],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":97,"show_sort_weight":98,"slug":99},"Exam",70,"exam",{"id":101,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},5,"Comic",60,"comic",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},6,"Technology",50,"technology",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":112,"show_sort_weight":113,"slug":114},7,"Healthcare",40,"healthcare",{"id":116,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":117,"slug":118},8,30,"research-report",{"id":120,"doc_module":4,"doc_module_name":25,"category_name":121,"show_sort_weight":122,"slug":123},9,"Religion & Spirituality",20,"religion-spirituality",{"id":122,"doc_module":4,"doc_module_name":25,"category_name":125,"show_sort_weight":122,"slug":126},"World Cup","world-cup",{"id":128,"doc_module":4,"doc_module_name":25,"category_name":129,"show_sort_weight":128,"slug":130},10,"Lifestyle","lifestyle",{"id":132,"doc_module":4,"doc_module_name":25,"category_name":133,"show_sort_weight":101,"slug":134},19,"General","general",{"code":4,"msg":86,"data":136},{"doc_id":83,"user_id":137,"nickname":42,"user_avatar":138,"doc_module":4,"category_id":116,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":139,"file_id":140,"file_url":141,"file_type":142,"file_size":143,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":22,"language":144,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":145,"faqs":146,"seo_title":147,"seo_description":12,"update_tm":84,"read_time":30},962084928432,"https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","A Machine Learning Model of Perturb-  \n1, 2, 3  \nLiam Johnson, James  \nTrials  \nFinal Testing Accuracy for Each Dataset  \nK562 Whole Genome  \nK562 Essential  \nRPE1  \nKey Terms:  \nObjective  \nOverview of Hyperparameter Tuning  \nGoal: To train a machine learning model on gene expression perturbational datasets in order to predict perturbations in new datasets. Such a model would be useful in identifying the specific genetic origins of disturbance, or perturbations, in systems which exhibit largely altered expression profiles from blanket effects such as spaceflight.  \nPerturbation – A disturbance of a specific gene or genes due to an external interference.  \nTranscriptional Response – A shift in the expression profile of genes affected by an upstream perturbation.  \nPerturb-Seq Dataset – Large scRNA-Seq datasets produced from high-throughput CRISPR-CAS driven gene knockout and screening experiments conducted primarily in human and mouse cell lines  \n83.5% 84.0% 86.9%  \nCasaletto  \n, Lauren Sanders, Sylvain  \n-  \nCostes  \nNASA Ames Research Center, Blue Marble Space. 4  \n.  \n-  \nSeq Data for use in Space Flight  \nNASA Ames Research Center.  \nGene Expression Profile Analysis  \nMLP Optimization  \nMethods  \nProcess Workflow  \nNew RNA Sample  \nPerturb-Seq AnnData Object  \nCluster and Annotate  \nLeiden Clusters  \nMachine Learning Algorithm  \nLabels  \nTraining Data  \nProcess Overview:  \n• Selection of Perturb-Seq dataset  \n• 2022 Genome-Wide Perturb-Seq (GWPS) from MIT.  \nPerturbational Model  \nClusters  \nAssociated Perturbed Genes  \nPerturbational  \nSignature  \n• Pre-processing, filtering, and normalization  \n• Leiden clustering and data pipeline establishment  \n• Annotation of biological significance for each cluster using gene set enrichment analysis (GSEA)  \n• Selection of suitable machine learning model  \n• Scikit-Learn Multi-Layer Perceptron (MLP)  \n• Establishment and configuration of MLP  \n• Parameter and Hyperparameter optimization trials  \n• Refinement and selection of Perturbational Model  \n• Trial RNA Dataset  \n• OSD-91 simulated microgravity in human lymphoblastoid cells  \n• Prediction pipeline establishment  \n• Quality assessment of output and examination of biological significance  \nGWPS Dataset Components:  \nK562 Whole Genome Set  \n• K562 chronic myeloid leukemia cancer cell line  \n• 1,989,578 cells  \n• 9,854 knocked out genes  \n• 8,248 Tracked transcriptional signals  \nK562 Essential Genes Set  \n• K562 chronic myeloid leukemia cancer cell line  \n• 310,385 cells  \n• 2,058 knocked out genes  \n• 8,563 Tracked transcriptional signals  \nRPE1 Set  \n• Retinal pigment epithelial non-cancerous cell line  \n• 247,914 cells  \n• 2,392 knocked out genes  \n• 8,749 Tracked transcriptional signals  \nResults and Discussion  \nThe OSD-91 RNA samples were successfully predicted as belonging to a single cluster in both the K562 Essential and RPE1 Perturb-Seq datasets.  \nK562 Essential  \nOSD-91Avg Sample Match: Cluster 18, 93.0%  \nMSigDB GSEA of the most prevalent target genes in cluster 18 reveals a primary association with construction and regulation of the DNA replication fork assembly.  \nRPE1  \nOSD-91Avg Sample Match:  \nCluster 3, 87.3%  \nMSigDB GSEA of the most prevalent target genes in cluster 3 reveals a primary association with the cytoskeletal structures and regulation of sister chromatid segregation, as well as a distinct marker of abnormal mitochondrial morphology  \nThe next step is to perform predictions using other sample RNA datasets. However, the creation of a truly genome-wide perturb-seq dataset with 20,000 targets and 20,000 tracked signals created in a non-cancerous cell line would be an invaluable tool.  \nAcknowledgements:  \nReplogle , J. M. , Saunders, R. A. , Pogson , A. N. , Hussmann, J. A. , Lenail, A. , Guna, A. , Mascibroda , L. , Wagner, E. J. , Adelman, K. , Lithwick-Yanai , G. , Iremadze , N. , Oberstrass, F. , Lipson, D. , Bonnar, J. L. , Jost, M. , Norman, T. M. , & Weissman, J. S. (2022) . Mapping informat","cbCaikaVLbkxBuYb","https://ap.wps.com/l/cbCaikaVLbkxBuYb","pdf",1609867,"English","# Objective\n## Overview of Hyperparameter Tuning\n# Process Workflow\n## Dataset Selection and Pre-processing\n## Clustering and Annotation (Leiden, GSEA)\n## Model Setup (MLP) and Optimization Trials\n# GWPS Dataset Components\n## K562 Whole Genome Set\n## K562 Essential Genes Set\n## RPE1 Set\n# Results and Discussion\n## OSD-91 Predictions and Biological Associations\n# Acknowledgements","[{\"question\":\"What is the goal of the machine learning model in this work?\",\"answer\":\"To train a machine learning model on gene expression perturbational datasets so it can predict perturbations in new datasets, helping identify genetic sources of altered expression profiles such as those from spaceflight.\"},{\"question\":\"Which steps are included in the processing workflow before model training?\",\"answer\":\"The workflow selects a Perturb-Seq dataset, performs pre-processing, filtering, and normalization, applies Leiden clustering, and uses gene set enrichment analysis (GSEA) to annotate biological significance for each cluster.\"},{\"question\":\"What model and optimization approach are used?\",\"answer\":\"A scikit-learn Multi-Layer Perceptron (MLP) is selected, configured, and optimized via parameter and hyperparameter tuning trials, followed by refinement and selection of the perturbational model.\"},{\"question\":\"What do the results show for the OSD-91 RNA samples?\",\"answer\":\"OSD-91 RNA samples are predicted to belong to single clusters in both the K562 Essential and RPE1 Perturb-Seq datasets, with average sample matches of 93.0% (Cluster 18) and 87.3% (Cluster 3), respectively, supported by GSEA associations.\"}]","A Machine Learning Model of Perturb-Seq Data - Trials - Final Testing Accuracy for Each Dataset | PDF"]