Lcms peak intensity filtering

lcms-peak-intensity-filtering

Summary

Filter LC/MS-detected peaks by minimum intensity threshold to enrich signal-to-noise ratio and reduce false positives in stable isotope labelling detection workflows. This skill applies absolute intensity cutoffs to putative incorporations and base-peak isotopologue candidates downstream of peak picking and alignment.

When to use

Apply this skill after XCMS peak picking, alignment, and grouping when you have identified putative incorporations (via PuInc_seeker) or base-peak isotopologue candidates and need to exclude low-intensity peaks that are likely noise or instrument artifacts. Use it when your LC/MS dataset exhibits variable background signal and you want to focus on metabolites with robust signal strength relative to your instrument's detection limit.

When NOT to use

• Input is already a curated metabolite reference database or in silico spectral library (peaks are not from your experimental LC/MS run)
• You are performing untargeted discovery on samples with extremely low metabolite abundance where even weak peaks may represent true biology
• Intensity values are already normalized (e.g., by total ion current or internal standard) and do not reflect absolute detector signal

Inputs

• XCMS feature table (XCMSet object) after peak picking, alignment, and grouping
• PuInc_seeker output object (putative incorporations with fold-change and p-value annotations)
• m/z values, retention times, and intensity measurements from aligned LC/MS runs

Outputs

• Filtered PuInc_seeker result object with low-intensity putative incorporations excluded
• basepeak_finder result object (geoRge object) containing base-peak isotopologue pairs meeting intensity thresholds
• Subset of features with m/z, intensity, and isotope-pair annotations above threshold

How to apply

Establish two intensity thresholds tuned to your LC/MS platform and sample preparation: (1) a putative incorporation intensity limit (e.g., 4000 counts) applied during PuInc_seeker to filter candidate labeled/unlabeled isotopologues, and (2) a base-peak minimum intensity threshold (e.g., 2000 counts) applied during basepeak_finder to validate that identified base peaks meet minimum signal abundance. Both thresholds should be derived from your instrument's noise floor and blank samples; set them conservatively to retain true metabolite signal while eliminating instrumental noise. The rationale is that low-intensity peaks are prone to mass accuracy drift and unstable isotope-pair detection, making them unreliable for downstream identification and quantification.

Related tools

• geoRge (Container package providing PuInc_seeker and basepeak_finder functions that apply intensity filtering as integral workflow steps) — https://github.com/jcapelladesto/geoRge
• XCMS (Upstream peak picking, alignment, and grouping to generate the feature table on which intensity filtering is applied) — https://bioconductor.org/packages/release/bioc/html/xcms.html
• R (Execution environment for geoRge intensity filtering functions)

Examples

s2 <- basepeak_finder(PuIncR = s1, XCMSet = mtbls213, UL.atomM=12.0, L.atomM=13.003355, ppm.s=6.5, Basepeak.minInt=2000)

Evaluation signals

• Filtered output contains no peaks with intensity below the specified threshold (e.g., no base peaks < 2000 counts)
• Number of retained features is reduced by a reasonable proportion (typically 30–60%) relative to unfiltered input, indicating effective noise suppression without over-filtering
• Retained base-peak isotopologue pairs maintain consistent mass accuracy (within 6.5 ppm tolerance) and intensity ratios expected for labeled/unlabeled carbon-13 pairs
• Downstream database_query hits and metabolite identifications show improved specificity (fewer false positives) compared to unfiltered runs
• m/z and intensity distributions of filtered peaks align with expected metabolite mass ranges and instrument sensitivity

Limitations

• Intensity thresholds are dataset and instrument-specific; thresholds derived from mtbls213 may not transfer to different LC/MS platforms, ionization modes, or sample matrices without reoptimization
• Filtering by absolute intensity assumes uniform detector response across m/z range; non-linear or m/z-dependent sensitivity bias may cause systematic loss of low-abundance metabolites in certain mass windows
• Threshold selection requires prior knowledge of noise floor; if blank/negative control intensities are not characterized beforehand, threshold setting may be arbitrary or too stringent/permissive
• geoRge 1.0 reduced function arguments for usability; verify that your version's default or user-specified intensity parameters match your experimental design

Evidence

• [intro] basepeak_finder function to identify base peaks with specified mass accuracy and intensity thresholds: "basepeak_finder function to identify base peaks with specified mass accuracy and intensity thresholds"
• [methods] unlabelled atomic mass 12.0, labelled atomic mass 13.003355, mass accuracy tolerance 6.5 ppm, and base-peak minimum intensity threshold 2000: "unlabelled atomic mass 12.0, labelled atomic mass 13.003355, mass accuracy tolerance 6.5 ppm, and base-peak minimum intensity threshold 2000"
• [intro] Base peak minimum intensity of 2000: "Base peak minimum intensity of 2000"
• [intro] Putative incorporation intensity limit of 4000: "Putative incorporation intensity limit of 4000"
• [methods] s2 <- basepeak_finder(PuIncR=s1,XCMSet=mtbls213,ULtag="CELL_Glc12",Ltag="CELL_Glc13",sep.pos="f",UL.atomM=12.0,L.atomM=13.003355,ppm.s=6.5,Basepeak.minInt=2000): "s2 <- basepeak_finder(PuIncR=s1,XCMSet=mtbls213,ULtag="CELL_Glc12",Ltag="CELL_Glc13",sep.pos="f",UL.atomM=12.0,L.atomM=13.003355,ppm.s=6.5,Basepeak.minInt=2000)"
• [readme] Same functions, less arguments. I have updated the functions so they are less tiring to use.: "Same functions, less arguments. I have updated the functions so they are less tiring to use."