Logo EESSAQ Microbiome
| EESSAQ Microbiome
AboutRegisterTutorialContact

Partners

Made by

Logo IRDA

In collaboration with

Logo ADLAB
Logo Université Laval

With the financial support of

Logo MAPAQ

© 2026 EESSAQ Microbiome. All rights reserved.

An initiative of the Research and Development Institute for the Agri-Environment.

All tutorials

Academic dashboard guide

The academic dashboard is built for in-depth exploration of metagenomic datasets. It lets you analyse the structure of microbial communities and identify the environmental factors driving how they assemble.

6 sections · about 10 min read

Open the dashboard

On this page

  1. 1.Alpha diversity: local ecology
  2. 2.Correlation analyses: crossing numerical variables
  3. 3.Beta diversity: community structure
  4. 4.The spatial projection (UMAP)
  5. 5.Setting up environmental gradients
  6. 6.Interpreting the per-taxon visualizations

1Alpha diversity: local ecology

Before comparing samples against each other, it's worth defining what is measured within a single sample. Alpha diversity assesses richness — the number of variants or OTUs/ASVs — and evenness, that is, how their abundances are distributed, within one local environment.

Available metrics

Shannon index
Combines richness and evenness into a single value. It is the most commonly reported metric.
Simpson index
Particularly sensitive to dominant species: useful when the question concerns how concentrated abundance is.
Chao1
An estimate of pure richness, corrected for rare taxa the sequencing effort did not detect.

Recommended analytical approach

Favour a linear model to rigorously assess the effect of continuous environmental variables — pH, carbon, nitrogen — on these indices. Group comparisons via box plots, for example between cropping systems, then serve to illustrate those effects rather than to establish them.

2Correlation analyses: crossing numerical variables

Before projecting whole communities, you often need to understand how the metadata interact with one another and with the microbiome's global indices. The interface offers two levels of analysis.

123
  1. 1The filters: site type, cropping system, soil series group, parent material
  2. 2The correlation between two numerical variables
  3. 3The multiple correlation matrix
Academic dashboard interface: filter panel on the left, two-variable correlation chart in the centre with its biological variable dropdown, and multiple correlation matrix on the right.
Simple (bivariate) correlations
Visualize the relationship between a microbiome-derived variable — Shannon index, microbial respiration, nitrogen fixation — and a continuous environmental variable such as pH or clay percentage. The chart displays the regression line along with the associated coefficient and p-value.
Multiple correlations
Correlation matrices are essential for spotting collinearity among multiple physicochemical variables before building more complex statistical models. You choose which variables to include, the statistic displayed, and the axis ordering.

Two useful settings

The "Taille des points" slider opens up a dense scatter, and the "Variable catégorique" setting colours points by group — parent material, for instance — to reveal a relationship that only holds within a subset.

3Beta diversity: community structure

Beta diversity measures the distance — the compositional dissimilarity — between different samples. Where alpha diversity describes a sample in isolation, beta diversity describes what separates two samples.

Distance matrices

Bray-Curtis
Accounts for the relative abundance of each taxon: two samples sharing the same species but in very different proportions will be treated as dissimilar.
Jaccard
Considers only the presence or absence of taxa, without weighting by abundance.
UniFrac
Incorporates phylogeny: two taxa close together on the tree contribute less to the distance than two distant ones.

Why an ordination?

A distance matrix spanning thousands of taxa cannot be read directly. A dimensionality reduction algorithm — UMAP or PCoA — is therefore applied to project those distances into an interpretable 2D or 3D space.

4The spatial projection (UMAP)

The interface offers a view of the composition of the three taxa under study through UMAP (Uniform Manifold Approximation and Projection), a non-linear dimensionality reduction technique. It projects the complexity of each sample's relative abundance profile onto two dimensions (UMAP 1 and UMAP 2), preserving the local and global topological structure of the data better than a classical PCoA.

How to read the distance

In these charts, the distance between two points reflects how dissimilar their microbial composition is: the closer two samples are, the more alike their communities.

12
  1. 1Choosing which physicochemical variable to project
  2. 2Choosing the colour palette
Three side-by-side UMAP projections — prokaryotes, fungi, and microeukaryotes — coloured along a pH gradient, with the physicochemical variable and palette selectors above.

5Setting up environmental gradients

The tool lets you overlay continuous metadata onto the UMAP projection to visually assess the effect of a gradient — pH, sand percentage, carbon percentage — on microbial composition.

Physicochemical variable
This dropdown selects the environmental variable projected onto your samples. By default, the application displays a master variable such as pH.
Colour selection
This setting changes the colour palette. The default option, "Spectre", offers high contrast that makes extreme values easier to read.

6Interpreting the per-taxon visualizations

The application generates three independent visualization panes at once, so you can compare how the different kingdoms respond ecologically to the same gradient.

Prokaryotes
Micro-organisms without a nucleus, mainly bacteria and archaea. They play a key role in the soil's biogeochemical cycles.
Fungi
Filamentous eukaryotic organisms that break down organic matter, form symbioses with plants, or can be pathogenic.
Microeukaryotes
Non-fungal eukaryotic micro-organisms, such as protists, that regulate microbial communities and nutrient cycles.

Reading a distribution pattern

The colour scale — say, blue for low pH through to red for high pH — lets you identify patterns. If the red points (alkaline pH) separate clearly from the blue points (acidic pH) along a UMAP axis, that strongly suggests pH exerts significant selective pressure on the structure of the community under study.

Comparing the three kingdoms

A sharp gradient among prokaryotes but a diffuse one among fungi indicates that the factor being tested does not structure the two kingdoms with equal force — that is a finding in its own right, not a shortcoming of the projection.

A question about your data?

These guides cover the essentials of the interface. For questions about interpreting your results or about data access, the EESSAQ team is here to help.

Contact usOpen the dashboard