Semisupervised adversarial neural networks for single-cell classification

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.

Automated summary

scNym is a semi-supervised adversarial neural network that can transfer cell identity annotations between different experiments by learning rich representations of cell identities from both labeled and unlabeled datasets. It shows superior performance in transferring annotations across experiments and can synthesize information from multiple datasets to improve accuracy. Additionally, scNym models are well calibrated, interpretable, and can be enhanced with saliency methods.