Heatmap Plot

The heat map (also called a clustergram) is a graphic representation of the unsupervised hierarchical clustering of target (gene) expression across all samples or biogroups in the study. In the plot, the targets and samples are arranged according to the similarity of their gene expression. The software uses either Pearson's correlation or Euclidean distance to calculate distances between samples and assays for hierarchical clustering based on their ΔCT values.

In the hierarchical clustering calculation, each sample/assay data point is represented as a node in the plot that is successively joined to nodes nearest to it by branches until all points are combined into a single cluster. The distance between the clusters (the inter-cluster distance) is defined by the linkage method, which the software can calculate three different ways (single, complete, or average). For more information on hierarchical clustering and linkage methods, see D'haeseleer P (2005) How does gene expression clustering work? Nat Biotechnol 23(12):1499-1501. .

Note: If you have configured your study with biogroups, the software displays the Biogroup/Sample switch that allows you to organize the results in the Heatmap Plot. If displayed, select Biogroup to group the results by biogroup or select Sample to compare the samples individually.

View Settings

The software provides multiple options for configuring the Heatmap Plot for review. To change the view settings, click View , then select the appropriate view options for the heatmap plot:

Option

Settings

Distance measure

Select the method to use for the sample-assay distance calculation for hierarchical clustering:

  • Pearson's Correlation — Select to use the Pearson product-moment correlation coefficient (Pearson's r) to perform the distance calculation.

  • Euclidean Distance — Select to use Euclidean (L2) distance to perform the distance calculation.

Note: The Euclidean Distance method of distance calculation is sensitive to scaling and differences in average expression level, whereas the Pearson's Correlation method is not.

Clustering method

Select the linkage method to use for the hierarchical clustering calculations:

  • Single Linkage — The distance between clusters is calculated as the shortest distance between any two members.

  • Complete Linkage — The distance between clusters is calculated as the largest distance between any two members. Complete linkage performs well on most gene expression data[1].

  • Average Linkage (unweighted pair-group method using arithmetic averages; UPGMA) — The distance between clusters is calculated as the averaged distance between any two members.

Map Type

Select the type of Heatmap in which you want the software to display the results of the hierarchical clustering calculations. For each map type, the ΔCT value of the neutral/middle expression level (mean or median) is set such that red indicates an increase with a ΔCT value below the middle level, and green indicates a decrease, with a ΔCT value above the middle level.

Note: The color indicators (Red/Green by default) are determined by the Color Scheme setting.

  • Global (CT or CT Plus) — The middle expression level is set as the median of all ΔCT values in the study by default. Any data point in the plot can be compared relative to all others.

  • Assay-Centric — For each assay, the middle expression level is set as the median of all ΔCT values from all samples for that assay. Data points for a given assay can only be compared relative to other data points for that assay.

  • Sample-Centric — For each sample, the middle expression level is set as the median of all ΔCT values from all assays for that sample. Data points for a given sample can only be compared relative to other data points for that sample.

Color Scheme

Select a color scheme for the software to display the results of the hierarchical clustering calculations: Red/Green, Red/ Blue, or Green/Orange color schemes.

1 See D'haeseleer P (2005) How does gene expression clustering work? Nat Biotechnol 23(12):1499-1501.

Heatmap tooltips

When mousing over the Heatmap Plot, the software displays in a tooltip window the ΔCT and ΔCT + global control mean (or global median if global normalization is selected in the analysis settings) for the given sample/target combination. The global control mean is the mean CT value of all selected endogenous controls in the study. The global median is the median value used if global normalization was in use. This value (global control mean or global median) is added on to the ΔCT to better approximate the original CT (a rough estimate of expression level) calculated for each sample and given assay before normalization.

  1. Sample name for the datapoint

  2. Target (assay) name for the datapoint

  3. ΔCT value that is calculated for the given sample-target combination

  4. ΔCT + global control mean calculated for the given sample-target combination.

Results Details table

Below the Heatmap Plot is the Results Details table, showing the following information:

Column

Use this column to

Sample

View the ID (a unique name or number) of the sample.

Biological Group

View the biological group (a unique name or number) to which the sample belongs.

Target

View the ID (a unique name or number) of the nucleic acid sequence targeted by the assay.

CT/CRT Mean

View the arithmetic average of the technical replicate Cq values.

Adjusted CT/ CRT Mean

View the average of the technical replicate Cq values that have been adjusted based on the "Maximum allowed CT" limit defined in the RQ Settings analysis settings.

Note: Wells with Cq scores greater than the "Maximum allowed CT" value are adjusted to the specified Cq limit.

∆CT/∆CRT Mean

View the arithmetic average of the technical replicate ∆Cq values for the sample replicate group.

Note: The ∆CT/∆CRT mean is calculated at the reaction plate level and represents the mean difference between the target ∆CT/∆CRT values and the endogenous control ∆CT/∆CRT values for all the technical replicates for that sample that are present on the plate.

∆CT/∆CRT SE

View the sample standard deviation of the sample replicate group level Cq values.

Note: The ∆CT/∆CRT SE value is calculated differently for multiplex and singleplex experiments. For multiplex experiments, the calculation is at the well level. For singleplex experiments, the calculation combines the plate-level Cq value variation between the target and the endogenous control.

∆CT/∆CRT + Control Median

View the arithmetic average of the technical replicate ∆CT/∆CRT values for the sample replicate group added to the control median.