SignacX

SignacX classifies cellular phenotypes from single-cell RNA-sequencing (scRNA-seq) data using neural networks trained on flow-sorted bulk gene expression from the Human Primary Cell Atlas to support precision-medicine applications such as identifying pathogenic cells and candidate drug targets.

Key Features:

• Neural network-based classification: Employs neural networks trained on flow-sorted bulk gene expression data from the Human Primary Cell Atlas to classify cellular phenotypes in scRNA-seq datasets.
• Cross-context robustness: Demonstrates classification performance across different diseases, sequencing technologies, species, and tissues.
• Conserved gene-signature detection: Reveals conserved gene signatures across conditions to standardize single-cell readouts.
• Autoimmune-disease targeting: Identifies cellular phenotypes relevant to autoimmune diseases to inform precision-medicine strategies aimed at reducing severe side effects and inducing immune tolerance.
• Drug-target discovery: Identifies known and novel immune-relevant candidate drug targets, including 12 potential precision-medicine targets reported for rheumatoid arthritis.

Scientific Applications:

• Cell type classification: Assigns cell-type and phenotype labels within scRNA-seq datasets to elucidate cellular diversity and function.
• Disease research: Identifies pathogenic cell populations in complex diseases such as autoimmune disorders to guide targeted therapeutic development.
• Cross-tissue and cross-species analysis: Enables comparative studies of cellular behavior across tissues and species by applying consistent classification criteria.
• Therapeutic target identification: Supports discovery of immune-relevant candidate drug targets for precision-medicine interventions.

Methodology:

SignacX employs neural networks trained on flow-sorted bulk gene expression data from the Human Primary Cell Atlas to classify cellular phenotypes in scRNA-seq datasets.