scFEA

scFEA predicts metabolic fluxes and metabolite variations from transcriptomics data to quantify metabolic heterogeneity at the single-cell and sample levels.

Key Features:

• Single-cell flux prediction: Estimates metabolic flux and metabolite variation at the single-cell level using transcriptomics data.
• Sample-level flux prediction: Aggregates predictions to provide flux estimates at broader sample levels in addition to single cells.
• Transcriptomics input: Uses transcriptomic profiles as the primary input to infer metabolic activity.
• Leverages transcriptomic–metabolomic correlation: Exploits established correlations between transcriptomic and metabolomic profiles to inform flux estimation.
• Unsupervised algorithm (single cell flux estimation analysis, scFEA): Employs an unsupervised analytical framework named single cell flux estimation analysis (scFEA).
• Neural network architecture: Utilizes a neural network architecture developed to estimate reaction rates from transcriptomics data.
• Cross-species support: Applicable to human, mouse, and 15 other common experimental model organisms.
• Scalability: Supports large-scale analyses of transcriptomics datasets.

Scientific Applications:

• Quantification of single-cell fluxome: Provides quantitative assessment of metabolic flux at single-cell resolution.
• Investigation of metabolic heterogeneity in disease: Enables analysis of metabolic variability and heterogeneity relevant to disease mechanisms.
• Comparative studies across species: Facilitates cross-species comparisons of metabolic flux using transcriptomics from humans, mice, and other model organisms.
• Exploration of metabolic dynamics: Supports study of metabolic dynamics and metabolite variation implications in health and disease.

Methodology:

Applies an unsupervised single cell flux estimation analysis (scFEA) that leverages transcriptomic–metabolomic correlations and a neural network architecture to estimate reaction rates and predict metabolic fluxes and metabolite variations from transcriptomics data at single-cell and sample levels.