Gene Expression Heatmap
Visualizza matrici di espressione genica con heatmap SVG interattiva. Z-score normalizzazione per riga, clustering gerarchico single-linkage, export PNG. Tool client-side: i dati non lasciano il browser.
Matrice CSV in input
Prima riga: intestazione (gene,condizione1,condizione2,...). Righe successive: nome gene + valori numerici separati da virgola. Max visualizzabili: 8 geni × 6 condizioni.
Heatmap
Mostra tabella dati (accessibilità)
| Gene | ctrl_1 | ctrl_2 | treat_1 | treat_2 |
|---|---|---|---|---|
| TP53 | 1.20 | 1.40 | 3.50 | 3.20 |
| BRCA1 | 2.10 | 2.30 | 1.10 | 0.90 |
| MYC | 0.50 | 0.70 | 2.80 | 3.00 |
| EGFR | 1.80 | 2.00 | 2.50 | 2.70 |
| PTEN | 2.50 | 2.30 | 0.80 | 0.60 |
Per studenti di Bioinformatica LM-91
Come interpretare la heatmap
- Colori caldi (rosso): espressione elevata o z-score positivo. Indica up-regulation rispetto alla media di riga.
- Colori freddi (blu): espressione bassa o z-score negativo. Indica down-regulation.
- Z-score: normalizza i valori di ogni gene per confrontare pattern tra geni con scale di espressione diverse. Formula: z = (x − μ) / σ, dove μ e σ sono media e deviazione standard della riga.
- Clustering single-linkage: raggruppa geni con profili di espressione simili usando la distanza euclidea minima tra cluster. Utile per identificare co-regolazione genica.
Applicazioni in ricerca
- Analisi differenziale di espressione (DEA) in RNA-seq
- Identificazione di pattern di co-espressione
- Visualizzazione risultati da DESeq2, edgeR, limma
- Analisi di heat shock proteins, oncogeni, tumor suppressors
I geni del dataset esempio (TP53, BRCA1, MYC, EGFR, PTEN) sono classici tumor suppressors e proto-oncogeni nel contesto di studi di espressione differenziale.
Questo tool fa parte della sezione Università e supporta studenti delle magistrali LM-91 in Bioinformatica:
- UniBa - primo corso magistrale Bioinformatica del Sud Italia
- Polimi + UniMi - Bioinformatics for Computational Genomics congiunto LM-91
Approfondisci la teoria con la serie UniAppunti / Bioinformatica.
Come utilizzare Gene Expression Heatmap
Attach or load the CSV matrix
Use the format gene,condition1,condition2,... in the first line, then a line for each gene with numeric values. You can start from the example dataset (5 genes x 4 conditions) with "Load Example".
Activate z-score and/or clustering
Normalization of z-scores per row makes genes comparable on different expression scales; hierarchical clustering using single linkage orders genes by profile similarity based on Euclidean distance.
Read the heatmap
Move the mouse (or keyboard focus) over a cell to see raw value and z-score in tooltip. Warm colors (red) indicate up-regulation, cool colors (blue) indicate down-regulation.
Export or view accessible table
Use "Export PNG" to download the heatmap image, or open "Show table data" for a screen reader-friendly table view.
Suggerimenti
- Use the z-score when comparing genotypes with vastly different expression ranges, not just between conditions.
- Activate clustering to visually identify groups of genes with similar co-regulation patterns.
- Always open the data table accessible if you need to verify exact values behind a color in the heatmap.
Domande frequenti
How is the row's z-score calculated?
For each gene (row), z = (x - mu) / sigma, where mu and sigma are the mean and standard deviation of that row's values. The result is limited to -3 to +3 on the color scale: 0 corresponds to white, positive values to red, negative values to blue.
What hierarchical clustering algorithm does it use?
The tool implements single-linkage clustering: starts with each gene as a separate cluster, calculates the Euclidean distance between expression profiles (using z-scores if active, otherwise raw values), and iteratively unites the two closest clusters, updating distances with the minimum of the merged clusters.
What is the maximum matrix size that can be displayed?
Heatmap shows up to 8 genes and 6 conditions at maximum for readability on mobile screens. If the CSV matrix exceeds these limits, only the first 8 genes and the first 6 conditions are displayed with a screen warning.
Are my genetic expression data being uploaded to a server?
No, all parsing of CSV, calculation of z-score, clustering and SVG rendering happen entirely in the browser. No data is sent to an external server, so you can use this tool on private educational datasets as well.
Can I use this tool for real RNA-seq analysis (DESeq2, edgeR, limma)?
It was primarily designed as a teaching tool for Bioinformatics courses (e.g. LM-91): you can stick together a subset of results already calculated by DESeq2, edgeR or limma to quickly visualize them, but it does not replace the full differential analysis software on real RNA-seq datasets.