TRC

TRC computes a truncated-rank correlation to quantify test-retest reliability between pairs of mass spectrometry (MS) experiments.

Key Features:

• Truncated top-rank correlation (top m-ranks): Calculates correlation using only the truncated top ranks of ranked peaks to emphasize high-abundance signals.
• Robustness to noise and high dimensionality: Designed to reduce the impact of experimental noise and masking of low-abundance compounds in high-dimensional MS datasets.
• MS peak data model: Operates on thousands of peaks characterized by mass-to-charge ratio (m/z), chromatographic retention time, and abundance measurements.
• Performance evaluation: Shown in a numerical study to outperform sample correlation and Kendall's τ for reliability assessment in noisy MS data.
• R package implementation: Implemented in the R package 'trc', which includes the TRC computation and related functions.

Scientific Applications:

• Test-retest reliability assessment: Quantifies similarity between repeated MS experiments to evaluate consistency across measurements.
• Biological replicate analysis: Applied to metabolome replicate analysis in HEK293 cells to assess reproducibility.
• Clinical metabolomic profiling: Applied in metabolomic profiling studies of benign prostate hyperplasia (BPH) patients to evaluate measurement reliability.

Methodology:

Correlation is computed using truncated top ranks (top m-ranks) of ranked MS peaks; performance was evaluated via numerical comparison to sample correlation and Kendall's τ; methods are implemented in the R package 'trc'.