AGENT SKILL

metabolomics-analysis

Untargeted and targeted metabolomics data analysis using pyMZML, mzTab-M processing, and metabolite identification. Use for LC-MS/GC-MS feature detection, metabolite annotation, differential metabolite analysis, pathway mapping, and metabolic flux visualization. Covers XCMS-compatible workflows in Python.

License 许可
MIT
最近更新
2026-06-05
Stars
★ 2
Forks
0

01 前言 Front Matter

name
metabolomics-analysis
license
MIT license
metadata
{ "skill-author": "Lord1Egypt" }
description
Untargeted and targeted metabolomics data analysis using pyMZML, mzTab-M processing, and metabolite identification. Use for LC-MS/GC-MS feature detection, metabolite annotation, differential metabolite analysis, pathway mapping, and metabolic flux visualization. Covers XCMS-compatible workflows in Python.
allowed-tools
Read Write Edit Bash

03 SKILL.md 正文

Metabolomics Analysis

Overview

Metabolomics is the large-scale study of small molecules (metabolites) within cells, tissues, or biofluids. This skill covers untargeted LC-MS and GC-MS metabolomics workflows including feature extraction, normalization, statistical analysis, metabolite identification, and pathway enrichment using Python-based tools.

When to Use This Skill

  • Processing LC-MS or GC-MS raw data for metabolite profiling
  • Feature detection and alignment across samples
  • Metabolite identification using MS2 matching (GNPS, mzCloud, HMDB)
  • Statistical analysis: PCA, PLS-DA, ANOVA, fold-change
  • Differential metabolite analysis between biological conditions
  • Metabolic pathway mapping (KEGG, BioCyc, MetaCyc)
  • Normalization methods for metabolomics data
  • Multi-omics integration with proteomics and transcriptomics

Quick Start

Loading mzML Data with pyMZML

from pyteomics import mzml
import numpy as np
import pandas as pd

# Load mzML file
spectra = []
with mzml.MzML("sample.mzML") as reader:
    for spectrum in reader:
        if spectrum["ms level"] == 1:
            spectra.append({
                "scan": spectrum["index"],
                "rt": spectrum["scanList"]["scan"][0]["scan start time"],
                "mz": spectrum["m/z array"],
                "intensity": spectrum["intensity array"],
            })

print(f"MS1 spectra loaded: {len(spectra)}")
print(f"RT range: {spectra[0]['rt']:.2f} - {spectra[-1]['rt']:.2f} min")

Feature Detection and Peak Picking

import numpy as np
import pandas as pd
from scipy.signal import find_peaks
from scipy.ndimage import gaussian_filter1d

def extract_ion_chromatogram(spectra: list, mz_target: float, ppm: float = 10.0) -> pd.DataFrame:
    """Extract Extracted Ion Chromatogram (EIC) for a target m/z."""
    mz_tol = mz_target * ppm / 1e6
    eic = []
    for s in spectra:
        mask = np.abs(s["mz"] - mz_target) <= mz_tol
        intensity = s["intensity"][mask].sum() if mask.any() else 0.0
        eic.append({"rt": s["rt"], "intensity": intensity})
    return pd.DataFrame(eic)

def detect_peaks(eic: pd.DataFrame, min_intensity: float = 1000.0) -> list:
    """Detect chromatographic peaks in EIC."""
    smoothed = gaussian_filter1d(eic["intensity"].values, sigma=2)
    peaks, props = find_peaks(smoothed, height=min_intensity, prominence=500, width=3)
    return [
        {
            "rt": eic["rt"].iloc[p],
            "peak_intensity": smoothed[p],
            "width": props["widths"][i],
        }
        for i, p in enumerate(peaks)
    ]

# Example: extract glucose (M+H = 181.0708 Da)
eic = extract_ion_chromatogram(spectra, mz_target=181.0708, ppm=5.0)
peaks = detect_peaks(eic, min_intensity=5000.0)
print(f"Peaks found for m/z 181.0708: {len(peaks)}")
for p in peaks:
    print(f"  RT={p['rt']:.2f} min, Intensity={p['peak_intensity']:.0f}")

HMDB Metabolite Search

import requests
import xml.etree.ElementTree as ET

def search_hmdb_by_mz(mz: float, ppm_tolerance: float = 10.0) -> list:
    """Search HMDB for metabolites matching a given m/z value."""
    ppm_tol_da = mz * ppm_tolerance / 1e6
    url = f"https://hmdb.ca/metabolites/search?query={mz:.4f}&search_type=chemical_formula"
    # HMDB REST API (mass search)
    url = f"https://hmdb.ca/metabolites.xml?monisotopic_molecular_weight={mz:.4f}&ppm={int(ppm_tolerance)}"
    try:
        r = requests.get(url, timeout=30)
        root = ET.fromstring(r.text)
        hits = []
        for metabolite in root.findall(".//metabolite"):
            name = metabolite.findtext("name", "")
            accession = metabolite.findtext("accession", "")
            formula = metabolite.findtext("chemical_formula", "")
            mono_mw = metabolite.findtext("monisotopic_molecular_weight", "")
            hits.append({
                "name": name,
                "hmdb_id": accession,
                "formula": formula,
                "mono_mw": mono_mw,
            })
        return hits
    except Exception as e:
        return [{"error": str(e)}]

hits = search_hmdb_by_mz(181.0708, ppm_tolerance=5.0)
for h in hits[:5]:
    print(f"  {h['name']} ({h['hmdb_id']}) — {h['formula']}")

Feature Matrix Normalization

import pandas as pd
import numpy as np

def normalize_metabolomics(df: pd.DataFrame, method: str = "pqn") -> pd.DataFrame:
    """
    Normalize metabolomics feature matrix.
    methods: 'pqn' (probabilistic quotient), 'sum', 'median', 'zscore'
    Rows = samples, Columns = features
    """
    if method == "pqn":
        # Probabilistic Quotient Normalization
        reference = df.median(axis=0)
        quotients = df.divide(reference, axis=1)
        correction_factors = quotients.median(axis=1)
        return df.divide(correction_factors, axis=0)
    elif method == "sum":
        return df.divide(df.sum(axis=1), axis=0) * df.sum(axis=1).median()
    elif method == "median":
        return df.divide(df.median(axis=1), axis=0) * df.median(axis=1).median()
    elif method == "zscore":
        return (df - df.mean()) / df.std()
    else:
        raise ValueError(f"Unknown method: {method}")

# Log transform and normalize
feature_matrix = pd.read_csv("features.csv", index_col=0)
log_matrix = np.log2(feature_matrix.replace(0, np.nan).fillna(feature_matrix.min().min() / 2))
normalized = normalize_metabolomics(log_matrix, method="pqn")

PCA and PLS-DA

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import PCA
from sklearn.cross_decomposition import PLSRegression
from sklearn.model_selection import cross_val_score

def run_pca(feature_matrix: pd.DataFrame, groups: pd.Series, title: str = "PCA"):
    """PCA with sample groups colored."""
    scaler = StandardScaler()
    X_scaled = scaler.fit_transform(feature_matrix.fillna(0))
    pca = PCA(n_components=2)
    coords = pca.fit_transform(X_scaled)

    fig, ax = plt.subplots(figsize=(8, 6))
    unique_groups = groups.unique()
    colors = plt.cm.Set1(np.linspace(0, 1, len(unique_groups)))

    for group, color in zip(unique_groups, colors):
        mask = groups == group
        ax.scatter(coords[mask, 0], coords[mask, 1], label=group, c=[color], s=80, alpha=0.8)

    ax.set_xlabel(f"PC1 ({pca.explained_variance_ratio_[0]*100:.1f}%)", fontsize=11)
    ax.set_ylabel(f"PC2 ({pca.explained_variance_ratio_[1]*100:.1f}%)", fontsize=11)
    ax.set_title(title, fontsize=13)
    ax.legend()
    plt.tight_layout()
    plt.savefig(f"{title.lower().replace(' ', '_')}.png", dpi=150)
    return pca, coords

groups = pd.Series(["Control"] * 6 + ["Treatment"] * 6, index=feature_matrix.index)
pca, coords = run_pca(normalized, groups, "Metabolomics PCA")

Differential Metabolite Analysis

import pandas as pd
import numpy as np
from scipy import stats
from statsmodels.stats.multitest import multipletests

def differential_metabolites(
    matrix: pd.DataFrame,
    group1: list,
    group2: list,
    fold_change_threshold: float = 1.5,
    fdr_threshold: float = 0.05,
) -> pd.DataFrame:
    """Mann-Whitney U test with FDR correction for metabolomics."""
    results = []
    for metabolite in matrix.columns:
        g1 = matrix.loc[group1, metabolite].dropna()
        g2 = matrix.loc[group2, metabolite].dropna()
        if len(g1) < 3 or len(g2) < 3:
            continue
        stat, pval = stats.mannwhitneyu(g1, g2, alternative="two-sided")
        fc = np.median(g2) - np.median(g1)  # Log2 FC
        results.append({"metabolite": metabolite, "log2FC": fc, "pvalue": pval})

    result_df = pd.DataFrame(results)
    _, fdr, _, _ = multipletests(result_df["pvalue"], method="fdr_bh")
    result_df["FDR"] = fdr
    result_df["significant"] = (
        (result_df["FDR"] < fdr_threshold) &
        (abs(result_df["log2FC"]) > np.log2(fold_change_threshold))
    )
    return result_df.sort_values("FDR")

ctrl_samples = [f"ctrl_{i}" for i in range(1, 7)]
treat_samples = [f"treat_{i}" for i in range(1, 7)]
diff_results = differential_metabolites(normalized.T, ctrl_samples, treat_samples)
print(f"Significant metabolites: {diff_results['significant'].sum()}")

KEGG Pathway Mapping

import requests
import pandas as pd

def get_kegg_pathways(hmdb_ids: list) -> pd.DataFrame:
    """Map metabolites to KEGG pathways using KEGG REST API."""
    pathway_data = []
    for hmdb_id in hmdb_ids[:50]:  # Respect rate limits
        try:
            # Convert HMDB ID to KEGG compound ID
            r = requests.get(f"https://hmdb.ca/metabolites/{hmdb_id}.xml", timeout=10)
            # Parse KEGG IDs from response
            import xml.etree.ElementTree as ET
            root = ET.fromstring(r.text)
            for pathway in root.findall(".//pathway"):
                pathway_data.append({
                    "hmdb_id": hmdb_id,
                    "pathway_name": pathway.findtext("name", ""),
                    "kegg_map_id": pathway.findtext("kegg_map_id", ""),
                })
        except Exception:
            continue
    return pd.DataFrame(pathway_data)

# Get significant metabolite HMDB IDs
sig_metabolites = diff_results[diff_results["significant"]]["metabolite"].tolist()
pathway_map = get_kegg_pathways(sig_metabolites[:20])
print("\nEnriched KEGG pathways:")
print(pathway_map["pathway_name"].value_counts().head(10).to_string())

Key Databases

Database Content URL
HMDB Human metabolites hmdb.ca
METLIN MS2 library metlin.scripps.edu
MassBank MS spectral library massbank.eu
GNPS Molecular networking gnps.ucsd.edu
KEGG Metabolic pathways kegg.jp
MetaboLights Public datasets ebi.ac.uk/metabolights

Dependencies

pip install pyteomics pandas numpy scipy matplotlib seaborn
pip install scikit-learn statsmodels
pip install requests lxml  # Database queries
pip install matchms  # Spectral similarity
pip install mzmine  # Feature detection (optional, Java-based)

04 同仓库更多 Skills

Lord1Egypt
Lord1Egypt/ai-skillforge

zarr-python

Chunked N-D arrays for cloud storage. Compressed arrays, parallel I/O, S3/GCS integration, NumPy/Dask/Xarray compatible, for large-scale scientific computing pipelines.

★ 2
Lord1Egypt
Lord1Egypt/ai-skillforge

what-if-oracle

Run structured What-If scenario analysis with multi-branch possibility exploration. Use this skill when the user asks speculative questions like "what if...", "what would happen if...", "what are the possibilities", "explore scenarios", "scenario analysis", "possibility space", "what could go wrong", "best case / worst case", "risk analysis", "contingency planning", "strategic options", or any question about uncertain futures. Also trigger when the user faces a fork-in-the-road decision, wants to stress-test an idea, or needs to think through consequences before committing.

★ 2
Lord1Egypt
Lord1Egypt/ai-skillforge

venue-templates

Access comprehensive LaTeX templates, formatting requirements, and submission guidelines for major scientific publication venues (Nature, Science, PLOS, IEEE, ACM), academic conferences (NeurIPS, ICML, CVPR, CHI), research posters, and grant proposals (NSF, NIH, DOE, DARPA). This skill should be used when preparing manuscripts for journal submission, conference papers, research posters, or grant proposals and need venue-specific formatting requirements and templates.

★ 2
Lord1Egypt
Lord1Egypt/ai-skillforge

vaex

Use this skill for processing and analyzing large tabular datasets (billions of rows) that exceed available RAM. Vaex excels at out-of-core DataFrame operations, lazy evaluation, fast aggregations, efficient visualization of big data, and machine learning on large datasets. Apply when users need to work with large CSV/HDF5/Arrow/Parquet files, perform fast statistics on massive datasets, create visualizations of big data, or build ML pipelines that do not fit in memory.

★ 2
Lord1Egypt
Lord1Egypt/ai-skillforge

usfiscaldata

Query the U.S. Treasury Fiscal Data API for federal financial data including national debt, government spending, revenue, interest rates, exchange rates, and savings bonds. Access 54 datasets and 182 data tables with no API key required. Use when working with U.S. federal fiscal data, national debt tracking (Debt to the Penny), Daily Treasury Statements, Monthly Treasury Statements, Treasury securities auctions, interest rates on Treasury securities, foreign exchange rates, savings bonds, or any U.S. government financial statistics.

★ 2
Lord1Egypt
Lord1Egypt/ai-skillforge

treatment-plans

Generate concise (3-4 page), focused medical treatment plans in LaTeX/PDF format for all clinical specialties. Supports general medical treatment, rehabilitation therapy, mental health care, chronic disease management, perioperative care, and pain management. Includes SMART goal frameworks, evidence-based interventions with minimal text citations, regulatory compliance (HIPAA), and professional formatting. Prioritizes brevity and clinical actionability.

★ 2