Genome Research · 2021 · 119 citations · 40 references
ComputingBioinformaticsComputer ScienceComputational BiologyGenomicsMachine LearningCell Type AnnotationDeep LearningMulti-task LearningSemi-supervised LearningSingle-cell AnalysisSingle-cell GenomicsAccuracyAnnotation TransferCell AnnotationCell Identity AnnotationGenerative Adversarial NetworksSaliency DetectionSemi-supervised AnnotationTransfer Annotation
Annotating cell identities is a common bottleneck in the analysis of single-cell genomics experiments. Here, we present scNym, a semisupervised, adversarial neural network that learns to transfer cell identity annotations from one experiment to another. scNym takes advantage of information in both labeled data sets and new, unlabeled data sets to learn rich representations of cell identity that enable effective annotation transfer. We show that scNym effectively transfers annotations across experiments despite biological and technical differences, achieving performance superior to existing methods. We also show that scNym models can synthesize information from multiple training and target data sets to improve performance. We show that in addition to high accuracy, scNym models are well calibrated and interpretable with saliency methods.
40
Comprehensive Integration of Single-Cell Data
Tim Stuart, Andrew Butler, Paul Hoffman et al. · Cell · 2019 · 16.9K citations · Full text
Fast, sensitive and accurate integration of single-cell data with Harmony
Ilya Korsunsky, Nghia Millard, Jean Fan et al. · Nature Methods · 2019 · 10.5K citations · Full text
A dendrite method for cluster analysis
T. Calinski, J. Harabasz · Communications in Statistics · 1974 · 6.7K citations