Principal component analysis
Upload a dataset to begin.
Start with your dataset
Upload a MetaboAnalyst-ready CSV. LipidLens will identify the groups automatically and let you choose what appears in the PCA.
Heatmap
Use the uploaded dataset and selected groups to generate a feature heatmap.
Generate a heatmap
Select groups on the left, then click Generate heatmap. The app will use the same feature filters and preprocessing settings as the PCA workflow.
Differential analysis
Identify experimental factors, confirm your biological question, and explore statistically tested lipid changes.
Confirm your experimental design
Factors come from your metadata or group labels. They are suggestions, never assumptions.
What do you want to investigate?
Your question determines the model and the post-hoc comparisons.
Requires exactly two complete observations per biological subject, one per level. It is not a substitute for a longitudinal mixed model.
Preprocessing and significance
Statistical inference uses log2 feature abundance; group differences are computed on the same scale.
Explore the results
Significance depends on the selected statistical effect or comparison.
Tested lipid features
Cross-comparison matrix
Selected lipid abundance
Select a feature in the table or volcano plot to inspect sample-level abundance and interaction patterns.
Download statistical results
Export significant hits, the full statistical table, the plan, or a self-contained project configuration.
Data formatter
Use a guided setup or map columns manually to generate an M/T sample table from any researcher spreadsheet.
Build your M/T file with a guided assistant
The assistant asks about sample identifiers, experimental conditions, QC, and naming. It uses your answers to generate the CSV and a group key. No research data leave your browser.
1. Upload a study file
CSV works offline. Excel files (.xlsx, .xls) are also supported when the SheetJS library loads.
2. Review your column mapping
3. Preview
4. Download outputs
The helper assigns sequential group numbers and preserves numeric feature columns. If QC samples are detected, they receive an additional numbered group. All original intensity values are retained.