scNovel

scNovel identifies novel rare cell types in single-cell RNA sequencing (scRNA-seq) data to enable discovery of previously uncharacterized cellular subtypes.

Key Features:

• Deep Learning-Based Neural Network: scNovel leverages a scalable deep learning neural network architecture tailored for detection of novel rare cells in single-cell RNA sequencing (scRNA-seq) datasets.
• High Performance: The method achieves an average Area Under the Receiver Operating Characteristic curve (AUROC) above 94%, representing up to a 16.26% improvement over the second-best method.
• Scalability: Validated on large-scale datasets, scNovel demonstrates the ability to handle datasets of varying scales and complexities efficiently.
• Robustness Across Datasets: The approach maintains effectiveness across diverse datasets generated by different protocols and exhibiting different degrees of class imbalance.

Scientific Applications:

• Biomedical Research: Facilitates discovery of novel cell types to provide new insights into cellular functions and disease mechanisms.
• Clinical Data Analysis: Applied to clinical datasets, including COVID-19 patient data where it identified three potential novel macrophage subtypes and detected consistent expression patterns of COVID-related differential genes.

Methodology:

scNovel employs a deep learning-based neural network architecture optimized for identification of rare cell types in scRNA-seq datasets and has been validated across diverse large-scale datasets.