M Z Alignment Across Samples

m-z-alignment-across-samples

Summary

Align mass tracks (EICs) detected in individual LC-MS samples into a unified MassGrid structure by leveraging high mass resolution to detect and resolve m/z differences corresponding to isotopes and adducts, then bin and centroid m/z values across the cohort. This enables reproducible tracking between detected features and their underlying chromatographic traces.

When to use

You have extracted mass tracks (EICs) from multiple LC-MS samples at 0.001 amu resolution and need to construct a sample-agnostic m/z reference frame. Apply this skill when you have ≥2 samples with overlapping mass tracks and systematic m/z drift between instruments or scan modes, or when you want to consolidate features across a cohort into a single feature table indexed by m/z.

When NOT to use

• Input mass tracks have already been merged or consolidated into a feature table—apply this skill during intermediate processing, not post-hoc.
• Mass resolution is <0.001 amu (e.g., low-resolution quadrupole or TOF instruments)—the skill depends on high mass resolution to prioritize mass separation.
• You are aligning only a single sample—the skill is designed for cross-sample alignment; intra-sample mass tracking uses different functions.

Inputs

• Mass tracks (EICs) extracted from each sample
• Binned m/z values at 0.001 amu resolution per sample
• Retention time-aligned chromatographic intensity traces
• Sample metadata (instrument, acquisition mode, optional reference sample designation)

Outputs

• MassGrid: unified m/z reference frame with one row per unique mass track
• MassGrid row content: m/z identifier, sample indices, sample-specific m/z values, anchor track flags
• Recalibration function (if systematic drift >1 ppm detected)
• Mass track-to-sample mapping (_mass_grid_mapping.csv)

How to apply

First, bin all m/z values within each sample at 0.001 amu resolution and merge adjacent bins within tolerance using nearest-neighbor clustering by m/z seeds to resolve overlapping m/z values (chromatograms.get_thousandth_bins, mass_functions.nn_cluster_by_mz_seeds). Identify anchor mass tracks in each sample by detecting m/z differences matching 13C/12C isotopes or Na/H adducts; select the reference sample as the one with the highest number of anchor tracks unless user-specified. For small studies (≤10 samples), perform pairwise alignment by comparing anchor mass tracks first and recalibrating all m/z values if systematic drift exceeds 1 ppm, then align remaining mass tracks. For larger studies (>10 samples), apply nearest-neighbor clustering on all mass tracks across samples to bin and centroid m/z values into a unified MassGrid. Validate that the MassGrid contains one m/z row per unique mass track identifier, each row lists sample indices and their corresponding m/z values, and anchor tracks are marked with higher priority.

Related tools

• pymzml (Parse mzML centroid spectra files to extract m/z and intensity data) — https://github.com/shuzhao-li/asari
• mass_functions.nn_cluster_by_mz_seeds (Perform nearest-neighbor clustering on m/z values by seeds to bin and resolve overlapping m/z) — https://github.com/shuzhao-li/asari
• chromatograms.get_thousandth_bins (Bin m/z values at 0.001 amu resolution and merge adjacent bins within tolerance) — https://github.com/shuzhao-li/asari
• chromatograms.extract_single_track_fullrt_length (Identify anchor mass tracks by detecting characteristic m/z differences (isotopes, adducts)) — https://github.com/shuzhao-li/asari
• MassGrid.build_grid_sample_wise (Perform pairwise sample-by-sample alignment for small studies (≤10 samples)) — https://github.com/shuzhao-li/asari
• MassGrid.build_grid_by_centroiding (Apply nearest-neighbor clustering for large studies to bin and centroid m/z across all samples) — https://github.com/shuzhao-li/asari
• MassGrid.bin_track_mzs (Bin sample-specific m/z values into unified MassGrid rows) — https://github.com/shuzhao-li/asari
• ext_Experiment.get_reference_sample_id (Identify or designate the reference sample with highest anchor mass track count) — https://github.com/shuzhao-li/asari

Examples

from asari.mass_functions import nn_cluster_by_mz_seeds; from asari.chromatograms import get_thousandth_bins; grid = MassGrid(samples, mass_tracks); grid.build_grid_by_centroiding() if len(samples) > 10 else grid.build_grid_sample_wise()

Evaluation signals

• MassGrid schema invariant: exactly one m/z row per unique mass track identifier; no duplicate rows.
• Each MassGrid row contains sample indices and corresponding m/z values for all samples contributing mass tracks to that m/z bin; samples not contributing that mass track are marked or absent.
• Anchor mass tracks are correctly flagged and have priority in alignment; verify by comparing 13C/12C and Na/H m/z offsets against theoretical values (1.003 amu for 13C/12C, 22.990 amu for Na/H adduct mass difference).
• Systematic m/z drift between samples (before recalibration) is ≤1 ppm post-alignment; drift >1 ppm triggers recalibration logic and should be corrected in output m/z values.
• Feature reproducibility: mass tracks within ±1 ppm of the MassGrid m/z row centroid are correctly backtracked to the same feature; inspect _mass_grid_mapping.csv for correct sample-to-feature assignments.

Limitations

• For studies >10 samples, centroiding-based alignment may reduce fine m/z precision if samples have large systematic drifts; recalibration is applied but assumes drift is uniform across m/z range.
• Anchor mass track identification depends on sufficient signal-to-noise and presence of isotope or adduct patterns; low-intensity or single-ion mass tracks may not have detectable anchors and alignment quality degrades.
• The 1 ppm drift threshold is empirically set for high-resolution Orbitrap data; different instrument types (Q-TOF, quadrupole) may require parameter adjustment.
• MassGrid does not retroactively align retention times; cross-sample retention time drift must be addressed in a separate alignment step before or after MassGrid construction.

Evidence

• [other] Asari aligns mass tracks across samples by constructing a MassGrid, a process that leverages high mass resolution to prioritize mass separation and alignment, enabling reproducible tracking and backtracking between features and mass tracks (EICs).: "Asari aligns mass tracks across samples by constructing a MassGrid, a process that leverages high mass resolution to prioritize mass separation and alignment, enabling reproducible tracking and"
• [other] Extract mass tracks from each sample using pymzml and bin m/z values at 0.001 amu resolution, merging adjacent bins within tolerance and applying nearest-neighbor clustering by m/z seeds to resolve overlapping m/z values: "Extract mass tracks from each sample using pymzml and bin m/z values at 0.001 amu resolution, merging adjacent bins within tolerance and applying nearest-neighbor clustering by m/z seeds to resolve"
• [other] Identify anchor mass tracks by detecting m/z differences matching 13C/12C isotopes or Na/H adducts within each sample: "Identify anchor mass tracks by detecting m/z differences matching 13C/12C isotopes or Na/H adducts within each sample"
• [other] For small studies (≤10 samples), perform pairwise alignment by comparing anchor mass tracks first, recalibrating all m/z values if systematic drift exceeds 1 ppm, then aligning remaining mass tracks: "For small studies (≤10 samples), perform pairwise alignment by comparing anchor mass tracks first, recalibrating all m/z values if systematic drift exceeds 1 ppm, then aligning remaining mass tracks"
• [other] For larger studies, apply nearest-neighbor clustering on all mass tracks across samples to bin and centroid m/z values into a unified MassGrid: "For larger studies, apply nearest-neighbor clustering on all mass tracks across samples to bin and centroid m/z values into a unified MassGrid"
• [other] MassGrid contains one m/z row per unique mass track identifier, each row lists sample indices and their corresponding m/z values, and anchor tracks are marked with higher priority in the alignment structure.: "MassGrid contains one m/z row per unique mass track identifier, each row lists sample indices and their corresponding m/z values, and anchor tracks are marked with higher priority in the alignment"
• [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"