GRM

GRM removes technical noise from single-cell RNA-seq data by leveraging ERCC (External RNA Controls Consortium) spike-in molecules to compute true gene expression levels for accurate transcriptome quantification.

Key Features:

• Technical noise reduction: Uses ERCC spike-in molecules to calibrate and remove technical variability in single-cell RNA-seq measurements.
• True gene expression computation: Explicitly computes corrected gene expression levels rather than focusing solely on differential expression detection.
• R implementation: Implemented in the R programming language for integration into R-based analysis workflows.

Scientific Applications:

• Clustering: Improves reliability of cell clustering by reducing technical variability in expression measurements.
• Trajectory inference: Enhances inference of developmental or differentiation trajectories through more accurate expression quantification.
• Differential expression analysis: Increases accuracy of differential expression results by normalizing technical biases with ERCC controls.
• Cellular heterogeneity studies: Facilitates detection of cellular heterogeneity and investigation of complex biological processes at single-cell resolution.

Methodology:

GRM benchmarks observed ERCC spike-in expression against known ERCC concentrations and normalizes single-cell RNA-seq measurements to correct technical biases across cells.