Mz Binning And Indexing

mz-binning-and-indexing

Summary

Organize high-resolution MS1 spectra into m/z-indexed data structures and merge adjacent bins to enable efficient mass track construction and alignment. This skill leverages integer-keyed indexing and proximity-based merging to separate and align masses prior to feature detection.

When to use

Apply this skill immediately after parsing mzML files into (m/z, scan_number, intensity) tuples when you need to build mass tracks from raw MS1 spectra. Use it when working with high-resolution instruments (e.g. Orbitrap) where mass separation and alignment must be prioritized over individual sample peak detection, and when you have ≥1 sample requiring reproducible mass track construction.

When NOT to use

• Input is already a feature table or pre-processed consensus feature matrix; binning is a preprocessing step, not a post-hoc refinement.
• Profile-mode (non-centroid) mzML files; binning assumes discrete m/z centroids, not continuous spectral profiles.
• Very low mass resolution data (e.g., nominal mass Orbitrap or nominal-mass quadrupole); the skill exploits high mass resolution to separate isobars and requires ≥5 ppm relative accuracy.

Inputs

• mzML file (centroid mode, MS1 spectra)
• List of (m/z, scan_number, intensity) tuples extracted from mzML
• Parameters: minimal required scan count per bin, absolute distance threshold (≤0.001 amu), ppm tolerance (instrument-specific)

Outputs

• mzTree dictionary: keys = int(mz × 1000), values = lists of (m/z, scan_number, intensity)
• Data bins: consolidated m/z intervals with merged adjacent bins
• Bin metadata: m/z range, number of contributing scans, scan count distribution

How to apply

Parse centroid mzML files using pymzml to extract all MS1 spectra as (m/z, scan_number, intensity) tuples. Index these points into an mzTree dictionary keyed by int(mz × 1000) for fast lookup. Apply get_thousandth_bins to create initial data bins from the mzTree, filtering bins that fail to meet a minimal required scan count threshold. Merge adjacent bins that are separated by ≤0.001 amu (0.001 Da) or fall within the ppm tolerance window (instrument-dependent, typically 5 ppm for Orbitrap). This two-stage merging—first by absolute distance, then by relative ppm—ensures that genuine isobars within the mass resolution are separated while noise and low-abundance spurious peaks are consolidated. The result is a set of consolidated mass bins, each representing a distinct m/z locus across the retention time range.

Related tools

• pymzml (Parse mzML files and extract MS1 spectra as (m/z, scan_number, intensity) tuples) — https://github.com/shuzhao-li/asari
• asari.chromatograms.get_thousandth_bins (Create initial data bins from mzTree using scan count filtering and adjacency-based merging) — https://github.com/shuzhao-li/asari
• asari.chromatograms.extract_massTracks_ (Construct mass tracks from consolidated bins for each sample) — https://github.com/shuzhao-li/asari

Examples

from asari.chromatograms import get_thousandth_bins; import pymzml; spectra = [(m, scan, intensity) for m, scan, intensity in pymzml.run('sample.mzML')]; mzTree = {int(m*1000): [] for m, s, i in spectra}; [mzTree[int(m*1000)].append((m, s, i)) for m, s, i in spectra]; bins = get_thousandth_bins(mzTree, min_scan_count=5, ppm_tolerance=5)

Evaluation signals

• Bins are non-overlapping and monotonically ordered by m/z; no bin spans >0.001 amu or exceeds ppm tolerance.
• Each bin contains ≥ minimum required scan count; low-abundance singleton m/z points are merged into adjacent bins.
• Mass separation is preserved: isobars separated by >ppm tolerance remain in distinct bins; spurious points within tolerance are consolidated.
• mzTree keys are integers in range [min(int(mz × 1000)), max(int(mz × 1000))]; lookup time is O(1) for a given m/z.
• Downstream mass track construction produces one track per bin (or multiple tracks per bin if the m/z range within a bin exceeds 2 × ppm tolerance, triggering nearest-neighbor clustering); track consensus m/z matches the bin's observed m/z distribution.

Limitations

• Binning is sensitive to the choice of minimal scan count threshold; too low a threshold will retain noise spikes; too high will merge genuine low-abundance peaks.
• The 0.001 amu and ppm-based merging rules assume Orbitrap-class high-resolution instruments; results on lower-resolution instruments (e.g., nominal-mass TOF) may be unpredictable.
• Isotopic fine structure (e.g., 13C satellites) may be binned separately or together depending on ppm tolerance; the article notes that anchor mass tracks are identified post-hoc to resolve isotopic relationships.
• Very high dynamic range (>1E8 intensity) may require rescaling before or after binning; the article mentions a preset ceiling of 1E8 for mass track intensity.

Evidence

• [other] Parse the mzML file using pymzml to retrieve all MS1 spectra as a list of (m/z, scan number, intensity) tuples.: "Parse the mzML file using pymzml to retrieve all MS1 spectra as a list of (m/z, scan_number, intensity) tuples."
• [other] Index data points into an mzTree dictionary keyed by int(mz × 1000) for efficient retrieval.: "Index data points into an mzTree dictionary keyed by int(mz × 1000) for efficient retrieval."
• [other] Create data bins from mzTree using get_thousandth_bins, filtering for minimal required scan count and merging adjacent bins separated by ≤0.001 amu or within ppm tolerance.: "Create data bins from mzTree using get_thousandth_bins, filtering for minimal required scan count and merging adjacent bins separated by ≤0.001 amu or within ppm tolerance."
• [intro] Taking advantage of high mass resolution to prioritize mass separation and alignment: "Taking advantage of high mass resolution to prioritize mass separation and alignment"
• [other] The default method uses pymzml to parse mzML files.: "The default method uses pymzml to parse mzML files."