Metabolite Intensity Alignment

---
name: metabolite-intensity-alignment
description: Use when you have raw lipidomic and metabolomic data files generated by the Multi-ABLE method (high-pressure liquid chromatography–mass spectrometry output) and need to prepare them for multivariate analysis to detect differential lipids and metabolites across biological samples (e.
license: CC-BY-4.0
metadata:
  edam_operation: http://edamontology.org/operation_3628
  edam_topics:
  - http://edamontology.org/topic_3370
  - http://edamontology.org/topic_0091
  - http://edamontology.org/topic_3172
  tools:
  - R
  - MultiABLER
  - ProteoMM
  - limma
  - tidyverse
  techniques:
  - LC-MS
derived_from:
- doi: 10.1016/j.isci.2023.106881
  title: MultiABLER
- doi: 10.1021/acs.analchem.9b01842
  title: ''
evidence_spans:
- MultiABLER is a set of R functions
- MultiABLER is a set of R functions forms a seamless workflow
claims: []
provenance:
  collection: https://w3id.org/holobiomicslab/asb-skill/collection/metabolomics/v2
  assembled_by: scripts/collect_metabolomics_collection.py
  sources:
  - build: coll_multiabler_cq
    doi: 10.1016/j.isci.2023.106881
    title: MultiABLER
  dedup_kept_from: coll_multiabler_cq
schema_version: 0.2.0
attribution:
  generator: AgenticScienceBuilder
  original_doi: 10.1016/j.isci.2023.106881
  all_source_dois:
  - 10.1016/j.isci.2023.106881
  - 10.1021/acs.analchem.9b01842
  zenodo_doi: 10.5281/zenodo.20794027
  curators: []
  promoter: Louis-Félix Nothias
  sponsor: CNRS & Université Côte d'Azur
---

# metabolite-intensity-alignment

## Summary

Normalize and align spectral features (m/z and retention time) across lipidomic and metabolomic samples to enable consistent feature detection and quantification in multivariate analysis. This step is critical for identifying differential lipids and metabolites when analyzing small-volume tissue extracts processed via high-pressure liquid chromatography–mass spectrometry (HPLC-MS).

## When to use

Apply this skill when you have raw lipidomic and metabolomic data files generated by the Multi-ABLE method (high-pressure liquid chromatography–mass spectrometry output) and need to prepare them for multivariate analysis to detect differential lipids and metabolites across biological samples (e.g., diseased vs. control tissue). Use it after loading raw data files and before running multivariate statistical functions.

## When NOT to use

- Input is already a normalized and aligned feature table (e.g., a pre-processed matrix from a different workflow or platform).
- Data were generated by a non-Multi-ABLE protocol or a different mass spectrometry instrument type, as MultiABLER functions are tuned for the Multi-ABLE method's instrument configuration.
- You do not have access to MultiABLER package or its dependencies (ProteoMM, limma, tidyverse) and cannot install them in your R environment.

## Inputs

- Raw lipidomic and metabolomic data files (generated by Multi-ABLE method / HPLC-MS)
- Sample metadata or experimental design file (to associate samples with groups)
- MultiABLER R function definitions (MultiABLER.r)

## Outputs

- Aligned and normalized feature intensity matrix (samples × features)
- Quality control plots (alignment verification, intensity distributions)
- Feature annotation table (m/z, retention time, ion form assignments)
- Statistical summary tables (per-feature normalization parameters)

## How to apply

Load raw lipidomic and metabolomic data files into the R environment via MultiABLER functions. Execute the integrated preprocessing functions provided by the MultiABLER package to normalize peak intensities and align spectral features (m/z values and retention times) across all samples in the dataset. The alignment step ensures that the same molecular ion detected across different samples is treated as a single feature rather than multiple features, enabling consistent quantification. The workflow normalizes intensities to account for instrument drift and sample-to-sample variation. After alignment and normalization, generate quality control plots to verify that features are correctly matched across samples and that intensity distributions are appropriately scaled. Proceed to multivariate analysis only after confirming alignment success via QC visualizations and statistical summaries.

## Related tools

- **MultiABLER** (Provides integrated R functions for normalization and alignment of spectral features across lipidomic and metabolomic samples) — https://github.com/holab-hku/MultiABLER
- **R** (Execution environment for loading, executing, and calling MultiABLER alignment and normalization functions)
- **ProteoMM** (Dependency package providing multivariate statistical methods and feature handling utilities) — https://www.bioconductor.org/packages/release/bioc/html/ProteoMM.html
- **limma** (Dependency package providing linear modeling and intensity normalization routines) — https://www.bioconductor.org/packages/release/bioc/html/limma.html
- **tidyverse** (Dependency package for data manipulation and visualization of alignment and QC outputs) — https://www.tidyverse.org

## Examples

```
devtools::install_github("holab-hku/MultiABLER", dependencies = TRUE); source("MultiABLER.r"); normalized_data <- align_and_normalize(raw_lipidomic_data, raw_metabolomic_data)
```

## Evaluation signals

- QC plots show consistent feature alignment across all samples with minimal scatter in m/z and retention time dimensions; features should cluster tightly around expected values.
- Normalized intensity distributions across samples are approximately symmetric or log-normal after transformation, with outlier samples identified for investigation.
- Feature annotation table contains no duplicate m/z–retention time pairs within alignment tolerance (e.g., co-eluting isomers correctly resolved or merged per protocol).
- Downstream multivariate analysis (e.g., differential lipid/metabolite detection) yields statistically significant features with biological interpretability (e.g., known lipid classes or metabolites associated with atherosclerosis phenotype).
- Before and after alignment plots or metrics demonstrate that inter-sample retention time drift and intensity variation have been reduced compared to raw data.

## Limitations

- MultiABLER is specifically designed for data generated by the Multi-ABLE method; application to other mass spectrometry protocols or instruments may yield suboptimal alignment due to differences in instrumental parameters (e.g., mass accuracy, scan rate, ionization efficiency).
- Alignment performance depends on the quality and consistency of raw data; severely degraded or outlier samples may fail to align properly or bias the normalization model.
- The tutorial and documentation assume R proficiency; users without prior R experience may encounter installation or execution issues with package dependencies (ProteoMM, limma, tidyverse).

## Evidence

- [intro] Execute the integrated preprocessing functions to normalize and align spectral features across samples.: "Execute the integrated preprocessing functions to normalize and align spectral features across samples."
- [readme] MultiABLER is a set of R functions forms a seamless workflow that supports integrative processing and analysis of lipidomic and metabolomic data generated by the Multi-ABLE method: "MultiABLER is a set of R functions forms a seamless workflow that supports integrative processing and analysis of lipidomic and metabolomic data generated by the Multi-ABLE method"
- [readme] install the following packges in R and run the funcions in MultiABLER.r: "install the following packges in R and run the funcions in MultiABLER.r"
- [intro] Generate output tables, quality control plots, and statistical summaries as produced by the function suite.: "Generate output tables, quality control plots, and statistical summaries as produced by the function suite."