Ms2 spectrum collection from data dependent acquisition

MS2 Spectrum Collection from Data-Dependent Acquisition

Summary

Extract and combine multiple MS2 fragmentation spectra associated with a chromatographic peak from DDA LC-MS/MS data to produce a consensus spectrum for compound annotation. This skill recovers clean precursor-fragment relationships that DDA naturally provides, enabling confident spectral matching against reference libraries.

When to use

You have DDA LC-MS/MS raw data (mzML format) with detected chromatographic peaks at a target m/z value and retention time window, and you need to build a high-confidence MS2 consensus spectrum for that peak to match against reference spectra (e.g., Metlin, GNPS). Use this when a single MS2 scan is insufficient or noisy, and multiple fragmentation spectra of the same precursor have been acquired across the chromatographic apex.

When NOT to use

• Input is DIA or SWATH data — those modalities fragment all ions within windows simultaneously, producing chimeric spectra with no clear precursor-fragment relationship; use DIA-specific peak detection methods instead.
• Only a single MS2 scan is available for the peak — consensus building requires multiple fragmentation spectra to be effective; a single spectrum should be used directly.
• You have already computed a feature table with grouped compound signals — this skill targets individual chromatographic peaks, not aggregated feature abundance matrices.

Inputs

• DDA LC-MS/MS raw data file (mzML format)
• Target m/z value and retention time range
• Tolerance parameters (ppm window, typically 20 ppm)

Outputs

• Consensus MS2 spectrum object (Spectra class)
• Precursor m/z and retention time metadata
• Combined peak list with fragment m/z and intensity values

How to apply

First, load the DDA raw data and filter to the retention time range containing your peak of interest using filterRt(). Extract all chromatographic peaks at your target m/z (±20 ppm tolerance) using chromPeaks() with mz and ppm parameters. Retrieve all MS2 fragmentation spectra (msLevel=2) associated with the detected peak(s) using chromPeakSpectra(). Combine the multiple MS2 spectra into a single consensus spectrum using combineSpectra() with FUN=combinePeaks, ppm=20, peaks='intersect', and minProp=0.8 (requiring 80% of peaks to be present across spectra to ensure robust fragment signals). The 'intersect' mode and high minProp threshold ensure that only reproducible fragments across multiple scans are retained, improving signal-to-noise and specificity for downstream spectral matching.

Related tools

• xcms (Core package providing chromPeaks(), chromPeakSpectra(), and filterRt() functions to detect and extract MS2 spectra from DDA data) — https://github.com/sneumann/xcms
• Spectra (Data container and manipulation backend; combineSpectra() is implemented in Spectra to merge multiple MS2 spectra into consensus) — https://github.com/RforMassSpectrometry/Spectra
• MsBackendMgf (Backend for reading reference MS2 spectra from Metlin/GNPS MGF files for downstream similarity scoring)
• MsFeatures (Optional downstream tool for feature grouping and compounding if extending analysis to multiple features) — https://github.com/RforMassSpectrometry/MsFeatures

Examples

dda_data_filt <- filterRt(dda_data, rt = c(230, 610)); peaks <- chromPeaks(dda_data_filt, mz = 304.1131, ppm = 20); spectra_ms2 <- chromPeakSpectra(dda_data_filt, msLevel = 2L); consensus <- combineSpectra(spectra_ms2, FUN = combinePeaks, ppm = 20, peaks = 'intersect', minProp = 0.8)

Evaluation signals

• The consensus spectrum contains only fragment m/z values present in ≥80% of input MS2 spectra (minProp=0.8 threshold enforced); spot-check a few fragments to confirm they appear in the majority of scans.
• Precursor m/z and retention time metadata are retained and match the target peak location (e.g., m/z 304.1131 ±20 ppm, rt ~230–610 s range).
• Consensus spectrum intensity values are merged (summed or averaged per combinePeaks logic) and show improved signal relative to any single input MS2 scan.
• When plotted as a mirror plot against reference spectra using plotSpectraMirror(), the consensus spectrum shows visual alignment with the correct reference (e.g., Fenamiphos vs. Flumazenil); cosine similarity scores should be ≥0.7 for a confident match.
• The combined spectrum has fewer noise peaks and higher fragment intensity for authentic fragments compared to raw single MS2 scans, indicating successful denoising.

Limitations

• DDA acquisition inherently covers only the top N most intense ions per cycle, leading to poor metabolite coverage and potential missed minor peaks.
• If chromatographic peaks are very narrow or few MS2 scans overlap the peak apex, consensus building may yield insufficient redundancy; minimum 2–3 MS2 scans per peak are recommended.
• The ppm tolerance (20 ppm) and minProp threshold (0.8) are empirically set and may need adjustment for very low-abundance or high-noise data; tighter thresholds reduce consensus quality but stricter thresholds may eliminate valid peaks.
• Reference library completeness and quality (e.g., Metlin, GNPS) affect annotation confidence; absent or poorly annotated compounds will not be reliably identified even with high-quality consensus spectra.

Evidence

• [intro] DDA method description and trade-offs: "In this method the top N most intense ions (m/z values) from a MS1 scan are selected for fragmentation in the next N scans before the cycle starts again. This method allows to generate clean MS2"
• [intro] chromPeakSpectra function for extracting MS2 data: "Extract MS2 spectra associated with chromatographic peaks for annotation"
• [intro] combineSpectra parameters and rationale: "Combine multiple MS2 spectra into consensus spectrum using combinePeaks function"
• [intro] Workflow step: consensus spectrum construction with specific parameters: "Combine multiple MS2 spectra into a single consensus spectrum using combineSpectra() with FUN=combinePeaks, ppm=20, peaks='intersect', minProp=0.8"
• [intro] Compounding rationale for feature grouping and annotation: "Compounding aims now at grouping such features presumably representing signal from the same originating compound to reduce data set complexity (and to aid in subsequent annotation steps)."
• [intro] Mirror plot and similarity scoring for spectrum matching: "Match experimental MS2 spectrum against reference database spectra using similarity metrics"