Skip to content

Core Concepts

Understanding the underlying concepts of MINT helps in optimizing workflows and interpreting results.

Regions of Interest (ROIs)

In MINT, a Region of Interest (ROI) is the retention-time interval used to inspect, optimize, and extract a target signal. In practice, the ROI is defined by the pair rt_min and rt_max, while rt stores the current target apex or reference RT inside that interval.

  • ROI bounds: rt_min and rt_max define the chromatogram segment that MINT treats as the candidate peak region.
  • Reference RT: rt is the target RT used for annotations, ROI bootstrapping, and later refinement.
  • Import fallback: If only rt is supplied, MINT creates a temporary bootstrap ROI (rt ± 5.0 s) so the target can be processed immediately.
  • Normalization: If imported RT values are given in minutes, MINT converts them to seconds before deriving or validating ROI bounds.
  • Refinement: In the Optimization tab, MINT updates provisional ROI bounds from observed chromatogram shape, apex position, and signal envelope.

Sparsification

Chromatogram data can be extremely dense, especially in high-resolution MS1 measurements. MINT uses Sparsification to reduce the number of data points while preserving the essential shape of the peak.

  • How it works: Instead of keeping every recorded intensity, MINT identifies transition points (e.g., scan times where the signal starts to rise, reaches a peak, or returns to baseline).
  • Benefits:
    • Dramatically reduces database size.
    • Speeds up plot rendering in the browser.
    • Minimizes memory usage during analysis.

LTTB Downsampling

For real-time visualization of "full-range" chromatograms (before RT slicing), MINT employs the Largest Triangle Three Buckets (LTTB) algorithm.

  • Purpose: LTTB effectively downsamples the data to a fixed number of points (e.g., 1000) for visual representation.
  • Visual Integrity: Unlike simple decimation, LTTB preserves visual features like local minima and maxima, ensuring that the "Shadow Plot" accurately reflects the raw data shape.

RT Alignment Logic

Retention Time (RT) alignment in MINT is designed for simplicity and robustness.

  • Median Apex Strategy: MINT identifies the peak apex within a user-defined ROI across multiple samples. It then calculates the median of these apex positions to use as a stable "Reference RT".
  • Authoritative Shifts: For each sample, the "Shift" is defined as Reference_RT - Sample_Apex_RT. These shifts are applied to the chromatogram traces in the Optimization view, effectively "snapping" them together.
  • Persistence: Unlike transient visualization-only shifts, MINT alignment state is authoritative and persisted in the database, ensuring that downstream processing uses the corrected ROI bounds.

Progressive Loading (Shadow Plots)

MINT prioritizes UI responsiveness through progressive loading:

  1. Step 1: Load a pre-calculated "Shadow Plot" (Envelope) using binned/downsampled data. This provides an immediate overview of the peak shape across all samples.
  2. Step 2: Load detailed, sparsified chromatogram data in the background.
  3. Step 3: Re-render with full interaction (manual optimization, legend filtering) once the detailed data is available.