# Spectral Feature Standardization

> Use when after peak-picking stage completes on centroided mzML or netCDF raw LC-MS data via any of the three wrapped algorithms (Centwave, FeatureFinderMetabo, ADAP), when you need to pass the detected features to downstream SLAW stages (alignment, isotope/adduct grouping, gap-filling, MS2.

- Skill: `holobiomicslab/spectral-feature-standardization` (Agent Skill)
- Install (CLI): `npx skillmds@latest add holobiomicslab/spectral-feature-standardization`
- Raw SKILL.md: https://api.skillmd.com/api/skills/holobiomicslab/spectral-feature-standardization/raw
- Safety review: PASS (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: CC-BY-4.0
- Author: HolobiomicsLab (https://skillmd.com/u/holobiomicslab)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/holobiomicslab/spectral-feature-standardization

---


# Spectral Feature Standardization

## Summary

Convert peak-picking output from heterogeneous algorithms (Centwave, FeatureFinderMetabo, ADAP) into a unified internal feature matrix representation with consistent m/z, retention time, and intensity columns. This standardization enables seamless downstream integration of algorithm-specific results into common alignment, grouping, and gap-filling workflows.

## When to use

After peak-picking stage completes on centroided mzML or netCDF raw LC-MS data via any of the three wrapped algorithms (Centwave, FeatureFinderMetabo, ADAP), when you need to pass the detected features to downstream SLAW stages (alignment, isotope/adduct grouping, gap-filling, MS2 extraction) that expect a unified schema regardless of which peak picker was invoked.

## When NOT to use

- Input is already a unified feature table from a prior standardization step — re-standardizing introduces redundancy and potential schema conflicts.
- Profile (non-centroided) mzML data — SLAW requires centroided input; standardization assumes peaks are already resolved.
- DIA-MS (data-independent acquisition) experiments — SLAW and its wrapped algorithms target DDA only; MS2 spectra from DIA will be skipped.

## Inputs

- Peak table from Centwave (algorithm-native format)
- Peak table from FeatureFinderMetabo (algorithm-native format)
- Peak table from ADAP (algorithm-native format)
- Algorithm metadata (algorithm name, version, parameter set used)

## Outputs

- Standardized feature matrix (m/z, retention time, intensity columns)
- Algorithm provenance metadata (algorithm name, execution context)
- Feature identifiers linked to algorithm-specific peak IDs

## How to apply

Receive the algorithm-specific peak table output from whichever peak-picking algorithm was routed based on the configuration key (Centwave, FeatureFinderMetabo, or ADAP). Extract the essential feature attributes: m/z (mass-to-charge ratio), retention time (RT), and intensity values. Map algorithm-native column names and units to the SLAW canonical schema (e.g., resolving differences in intensity normalization or RT scale). Assemble these into a standardized feature matrix where rows are features and columns are [m/z, RT, intensity, sample_id, algorithm_source]. This standardized format decouples downstream stages from algorithm-specific quirks and allows the same alignment, grouping, and gap-filling logic to operate identically on Centwave, FeatureFinderMetabo, or ADAP results.

## Related tools

- **Centwave** (Peak-picking algorithm whose output is standardized)
- **FeatureFinderMetabo** (Peak-picking algorithm whose output is standardized)
- **ADAP** (Peak-picking algorithm whose output is standardized)
- **SLAW** (Wrapper framework that invokes peak-picking, receives native output, and coordinates standardization before downstream stages) — https://github.com/zamboni-lab/SLAW

## Evaluation signals

- All features from the algorithm-native peak table are present in the standardized matrix with no loss or duplication.
- Every row in the standardized matrix has valid numeric values for m/z (positive), retention time (non-negative), and intensity (non-negative or zero).
- Column schema matches the SLAW canonical schema (m/z, RT, intensity, sample_id, algorithm_source) regardless of input algorithm.
- Downstream alignment step can consume the standardized matrix and produce consistent RT correction parameters across samples without requiring algorithm-specific post-processing.
- Algorithm metadata is preserved and retrievable, allowing users to trace which peak picker generated each feature.

## Limitations

- Standardization assumes all three algorithms (Centwave, FeatureFinderMetabo, ADAP) produce functionally equivalent peak definitions (m/z, RT, intensity); systematic differences in peak boundary detection or intensity normalization between algorithms are not harmonized beyond schema unification.
- mzML and netCDF format support is fixed; other raw MS formats (e.g., vendor-native .raw or .d) must first be converted to mzML/netCDF before reaching peak-picking and standardization stages.
- DDA-only scope: standardization is designed for DDA (data-dependent acquisition) experiments; DIA-MS2 spectra will be skipped and not integrated into the feature matrix.

## Evidence

- [other] Standardize the output peak table to a common internal format (e.g., feature matrix with m/z, retention time, intensity columns).: "Standardize the output peak table to a common internal format (e.g., feature matrix with m/z, retention time, intensity columns)."
- [other] SLAW wraps three independent peak picking algorithms—Centwave, FeatureFinderMetabo, and ADAP—enabling configurable selection among them for the peak-picking stage of untargeted LC-MS processing.: "SLAW wraps three independent peak picking algorithms—Centwave, FeatureFinderMetabo, and ADAP—enabling configurable selection among them"
- [readme] Wrapping of three main peak picking algorithms: Centwave, FeatureFinderMetabo, ADAP: "Wrapping of three main peak picking algorithms: Centwave, FeatureFinderMetabo, ADAP"
- [other] Return the standardized peak table and algorithm metadata to the caller for downstream SLAW stages (alignment, grouping, gap-filling).: "Return the standardized peak table and algorithm metadata to the caller for downstream SLAW stages (alignment, grouping, gap-filling)."
- [readme] All data must be centroided and of unique polarity. Centroided mzML can be obtained with ProteoWizard.: "All data must be centroided and of unique polarity. Centroided mzML can be obtained"