BIDCell: Biologically-informed self-supervised learning for segmentation of subcellular spatial transcriptomics data
Fu X, Lin Y, Lin D, Mechtersheimer D, Wang C, Ameen F, Ghazanfar S, Patrick E, Kim J, Yang J. BIDCell: Biologically-informed self-supervised learning for segmentation of subcellular spatial transcriptomics data. Nature Communications 2024, 15: 509. PMID: 38218939, PMCID: PMC10787788, DOI: 10.1038/s41467-023-44560-w.Peer-Reviewed Original ResearchMeSH KeywordsBenchmarkingErythrocytes, AbnormalGene Expression ProfilingHistocompatibility TestingSupervised Machine LearningConceptsGene expressionSingle-cell transcriptomic dataSpatial expression analysisMap of gene expressionSpatial mapping of gene expressionTranscriptome dataBiological discoveryExpression analysisTranscriptomic platformsOversized cellsPublic repositoriesCell morphologyState-of-the-art methodsSelf-supervised learningDeep learning-based frameworkState-of-the-artTissue typesLearning-based frameworkHigh-resolution spatial mappingCellsExpressionSignificant analytical challengeSegmentation performanceLoss functionRecent advances