metabolomics-de
Metabolomics differential analysis using univariate tests (t-test, FDR), multivariate methods (PCA, PLS-DA, OPLS-DA, sPLS-DA), Random Forest, and ROC analysis for biomarker discovery.
SKILL.mdSkills/metabolomics/metabolomics-de/SKILL.md
GitHub📈 Metabolomics Differential Analysis
Univariate and multivariate statistical analysis for identifying differentially abundant metabolites and biomarker discovery.
Core Capabilities
- Univariate testing: t-test, Wilcoxon, ANOVA with BH FDR correction
- Fold change: Log2FC calculation with volcano plot visualization
- PCA: Unsupervised exploration and outlier detection
- PLS-DA / sPLS-DA: Supervised classification with VIP scores and sparse feature selection
- OPLS-DA: Orthogonal PLS-DA with S-plot for predictive features
- Random Forest: Non-linear feature importance ranking
- ROC analysis: Diagnostic performance evaluation of candidate biomarkers
CLI Reference
python omicsclaw.py run met-diff --demo
python omicsclaw.py run met-diff --input <feature_table.csv> --output <dir>
Algorithm / Methodology
Univariate Analysis (R)
library(tidyverse)
data <- read.csv('normalized_data.csv', row.names = 1)
groups <- factor(read.csv('sample_info.csv')$group)
# T-test for each feature
ttest_results <- apply(data, 2, function(x) {
test <- t.test(x ~ groups)
c(pvalue = test$p.value,
fc = mean(x[groups == levels(groups)[2]]) - mean(x[groups == levels(groups)[1]]))
})
ttest_results <- as.data.frame(t(ttest_results))
ttest_results$fdr <- p.adjust(ttest_results$pvalue, method = 'BH')
sig_features <- ttest_results[ttest_results$fdr < 0.05, ]
Volcano Plot (R)
library(ggplot2)
results$significant <- results$fdr < 0.05 & abs(results$log2fc) > 1
ggplot(results, aes(x = log2fc, y = -log10(pvalue), color = significant)) +
geom_point(alpha = 0.6) +
scale_color_manual(values = c('gray', 'red')) +
geom_hline(yintercept = -log10(0.05), linetype = 'dashed') +
geom_vline(xintercept = c(-1, 1), linetype = 'dashed') +
labs(x = 'Log2 Fold Change', y = '-log10(p-value)') +
theme_bw()
PCA (R)
library(pcaMethods)
pca_result <- pca(data, nPcs = 5, method = 'ppca')
scores <- as.data.frame(scores(pca_result))
scores$group <- groups
ggplot(scores, aes(x = PC1, y = PC2, color = group)) +
geom_point(size = 3) +
stat_ellipse(level = 0.95) +
labs(x = paste0('PC1 (', round(pca_result@R2[1] * 100, 1), '%)'),
y = paste0('PC2 (', round(pca_result@R2[2] * 100, 1), '%)')) +
theme_bw()
PLS-DA with VIP Scores (mixOmics)
library(mixOmics)
plsda_result <- plsda(as.matrix(data), groups, ncomp = 3)
# Cross-validation
perf_plsda <- perf(plsda_result, validation = 'Mfold', folds = 5, nrepeat = 50)
ncomp_opt <- perf_plsda$choice.ncomp['BER', 'centroids.dist']
# Final model + VIP scores
final_plsda <- plsda(as.matrix(data), groups, ncomp = ncomp_opt)
plotIndiv(final_plsda, group = groups, ellipse = TRUE, legend = TRUE)
vip <- vip(final_plsda)
top_vip <- sort(vip[, ncomp_opt], decreasing = TRUE)[1:20]
sPLS-DA (Sparse — Feature Selection)
tune_splsda <- tune.splsda(as.matrix(data), groups, ncomp = 3,
validation = 'Mfold', folds = 5, nrepeat = 50,
test.keepX = c(5, 10, 20, 50, 100))
optimal_keepX <- tune_splsda$choice.keepX
splsda_result <- splsda(as.matrix(data), groups, ncomp = ncomp_opt, keepX = optimal_keepX)
selected_features <- selectVar(splsda_result, comp = 1)$name
OPLS-DA (ropls)
library(ropls)
oplsda <- opls(data, groups, predI = 1, orthoI = NA)
plot(oplsda, typeVc = 'x-score') # Scores plot
plot(oplsda, typeVc = 'x-loading') # S-plot
vip_scores <- getVipVn(oplsda)
top_vip <- sort(vip_scores, decreasing = TRUE)[1:20]
Random Forest Feature Importance
library(randomForest)
rf_model <- randomForest(x = data, y = groups, importance = TRUE, ntree = 500)
importance <- importance(rf_model)
top_features <- rownames(importance)[order(importance[, 'MeanDecreaseAccuracy'], decreasing = TRUE)[1:20]]
varImpPlot(rf_model, n.var = 20)
ROC Analysis (pROC)
library(pROC)
top_feature <- 'feature_123'
roc_result <- roc(groups, data[, top_feature])
plot(roc_result, main = paste('AUC =', round(auc(roc_result), 3)))
Heatmap
library(pheatmap)
top_features <- rownames(sig_features)[1:50]
annotation_row <- data.frame(Group = groups)
rownames(annotation_row) <- rownames(data)
pheatmap(t(data[, top_features]), annotation_col = annotation_row,
scale = 'row', clustering_method = 'ward.D2')
Parameters
| Parameter | Default | Description |
|---|---|---|
--method |
ttest |
Statistical test method |
--fdr-cutoff |
0.05 |
FDR significance threshold |
--fc-cutoff |
1.0 |
Log2FC threshold |
--multivariate |
pca |
pca, plsda, oplsda, splsda |
Why This Exists
- Without it: Complex metabolomic variance is hard to parse; minor batch effects can mimic true biological signal
- With it: Powerful multivariate (PLS-DA) and univariate testing models separate true biomarkers from noise
- Why OmicsClaw: Automates strict R-based multidimensional workflows with single Python calls
Workflow
- Calculate: Log transformations and scaling.
- Execute: Run PCA/OPLS-DA and extract variable importance (VIP) scores.
- Assess: Execute parallel univariate FDR-corrected models.
- Generate: Output candidate biomarker lists.
- Report: Synthesize VIP plots, score plots, and hierarchical heatmaps.
Example Queries
- "Run PLS-DA and find significant metabolites"
- "Perform differential analysis on this normalized metabolomics matrix"
Output Structure
output_directory/
├── report.md
├── result.json
├── statistics.csv
├── figures/
│ ├── plsda_scores.png
│ ├── vip_plot.png
│ └── volcano_plot.png
├── tables/
│ └── differential_results.csv
└── reproducibility/
├── commands.sh
├── requirements.txt
└── checksums.sha256
Safety
- Local-first: Strict offline processing without external upload.
- Disclaimer: Requires OmicsClaw reporting structures and disclaimers.
- Audit trail: Hyperparameters and operational flow states are logged fully.
Integration with Orchestrator
Trigger conditions:
- Automatically invoked dynamically based on tool metadata and user intent matching.
Chaining partners:
met-normalize— Upstream data scalingmet-annotate— Downstream identification of biomarkers
Version Compatibility
Reference examples tested with: mixOmics 6.24+, ropls 1.32+
Dependencies
Required: numpy, pandas, scipy, scikit-learn Optional: mixOmics (R), ropls (R), randomForest (R), pROC (R), pheatmap (R)
Citations
- mixOmics — Rohart et al., PLoS Computational Biology 2017
- ropls — Thévenot et al., Journal of Proteome Research 2015
- MetaboAnalyst — Pang et al., Nucleic Acids Research 2021
Related Skills
met-normalize— Data normalization before analysismet-pathway— Pathway enrichment of significant metabolitesxcms-preprocess— Feature extraction upstream