Xcms Grouping Result Interpretation

xcms-grouping-result-interpretation

Summary

Interpret and validate XCMS grouping results by detecting misaligned feature groups using ncGTW's misalignDetect() function with dual bandwidth parameters. This skill identifies alignment artifacts that arise when XCMS assumes a single warping function across all m/z bins, enabling targeted realignment of problematic features before downstream analysis.

When to use

Apply this skill when XCMS grouping has been performed on LC-MS data from studies with hundreds of samples or data acquisition periods longer than a week, where retention time drift structures are complex and the single-warping-function assumption is likely violated. Use specifically when you suspect misaligned feature groups may compromise peak-regrouping, peak-filling, or quantification accuracy.

When NOT to use

• XCMS grouping has been performed on small, short-duration studies (< 100 samples, < 1 week acquisition) where RT drift is minimal and single-warping assumptions are reasonable.
• You lack two independent XCMS grouping results with systematically different bandwidth parameters.
• The ppm tolerance parameter for misalignDetect() does not match the ppm threshold used in the original XCMS peak detection step.

Inputs

• XCMS grouping result object (large bandwidth parameter set)
• XCMS grouping result object (small bandwidth parameter set)
• ppm tolerance parameter (numeric, matching peak detection ppm)

Outputs

• excluGroups table (detected misaligned feature groups)
• p-value estimates for each feature (from higher-resolution alignment)

How to apply

Load two XCMS grouping results into R: one with a large bandwidth (bw) parameter set to the expected maximal RT drift, and one with a small bw parameter set to the RT sampling resolution. Call misalignDetect() from the ncGTW package, providing both grouping objects and a ppm tolerance parameter that matches the ppm value used during initial peak detection. The function estimates p-values for each feature under the null hypothesis of accurate alignment (higher resolution result) and identifies features with sufficiently small p-values and disjoint sample subsets as misaligned. Extract the returned excluGroups table, which lists all detected misaligned feature groups as candidates for realignment with ncGTW's reference-free pairwise alignment method.

Related tools

• ncGTW (Provides misalignDetect() function to identify misaligned feature groups from XCMS and realign them using neighbor-wise compound-specific graphical time warping) — https://github.com/ChiungTingWu/ncGTW
• xcms (Produces grouping results (with different bandwidth parameters) that serve as input to misalignDetect() for misalignment detection)
• R (Runtime environment for executing ncGTW and XCMS functions)

Examples

result <- misalignDetect(xcmsGrouping_largeBW, xcmsGrouping_smallBW, ppm=5)

Evaluation signals

• excluGroups table is non-empty and contains feature groups with p-values below the statistical significance threshold, indicating detected misalignments
• Detected misaligned features show disjoint sample subsets (not random distribution across all samples), confirming structured misalignment rather than noise
• Feature m/z values and retention times in excluGroups match entries in the original XCMS grouping objects, validating traceability
• After realignment with ncGTW on detected misaligned features, subsequent peak-regrouping or peak-filling operations yield improved quantification consistency or reduced artifactual peak intensity variance across the sample cohort

Limitations

• misalignDetect() requires careful parameterization: bandwidth values must be chosen to represent both maximal RT drift and sampling resolution; incorrect bw selection will either miss true misalignments or generate false positives.
• The ppm parameter must exactly match the peak detection ppm threshold used in the original XCMS analysis; mismatches will produce unreliable p-value estimates.
• Requires two separate XCMS grouping runs with different parameters, adding computational and storage overhead.
• The method assumes that misaligned features will exhibit sufficiently small p-values and disjoint sample subsets; features with gradual or diffuse misalignment patterns across many samples may not be detected.

Evidence

• [other] misalignDetect() input requirements and parameters: "misalignDetect() requires two XCMS grouping results with different bw values (one set to expected maximal RT drift, one to RT sampling resolution) and a ppm parameter matching the peak detection ppm"
• [intro] XCMS limitation that motivates misalignment detection: "ncGTW uses individualized warping functions for different compounds and assigns constraint edges on warping functions of neighboring samples. That is, ncGTW avoids the popular but not accurate"
• [intro] Problem statement: misaligned features from XCMS: "Due to the same warping function assumption or bad parameter settings, xcms may have some misaligned features, and there is a function in ncGTW to identify such misalignments"
• [readme] ncGTW misalignment detection algorithm overview: "ncGTW detects the misaligned features with two criterions. First, ncGTW algorithm estimates the p-value of each feature using higher resolution alignment result, where the p-value is given by the"
• [other] Workflow: extract and return excluGroups table: "Extract and return the excluGroups table containing the detected misaligned feature groups"