Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

Cosmos-Hub allows researchers to create virtual cohorts for microbiome analysis using their study metadata. Simply upload any categorical or continuous metadata from your study via the software’s standard template and create virtual cohorts to compare against each other.

Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

Cosmos-Hub allows researchers to create virtual cohorts for microbiome analysis using their study metadata. Simply upload any categorical or continuous metadata variables from your study via the software’s standard template and create virtual cohorts from a study or conduct a meta-analysis across multiple studies in just a few clicks.

Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

Single Study Analysis

The Cosmos-Hub dashboard allows users to easily navigate their studies, select the samples they'd like to include and conduct their downstream analysis using the statistics toolbox. By selecting folders advanced filtering, users can select samples via their folder paths but also sub-select samples based on metadata variables of their choosing such as 'number of reads', 'time point', 'treatment group' or any numerical or text metadata variable they've uploaded.

Don’t have data from additional studies? Use the Cosmos-Hub Atlas to incorporate new datasets into your study and increase the statistical power of your study.

Meta-Analysis*

One of the most powerful features of the Cosmos-Hub platform is the ability to create virtual cohorts using your study metadata, using advanced filtering. This allows users to conduct new studies, usng samples from multiple projects without having to collect new samples. Simply use tools like MaAsLin to control for confounders across your studies and start your virtual study!

Don’t have data from additional studies? Use the Cosmos-Hub Atlas to incorporate new datasets into your study and increase the statistical power of your study.

Univariate Statistics

The Cosmos-Hub platform provides a univariate statistics workflow that enables users to conduct customizable pairwise comparative analyses of up to 24 cohorts to reveal statistical testing figures and visualizations.

Users have the flexibility to switch between different databases and metrics that have been run with a click of the button, easing the interpretation of the data. The outputs are interactive and exportable in a number of different formats including CSV data or PNG/SVG/PDF image formats.

Finally, when it’s time to publish, you can access each comparative analysis you’ve created in the main study dashboard where you will be able to access the methods and which parameters were used so that you or a peer can always reproduce the results.

Matrix table

This functional table allows you to browse your samples, look for outliers, search for and pin organisms of interest, switch between different levels of taxonomy or hierarchy, as well as notice any trends amongst your samples.

Heatmap

This highly interactive heatmap allows users to manipulate the output in real time and export any configuration of their choosing. Users can change the way the heatmap is clustered, the taxonomy/hierarchy displayed, the number of features shown as well as the colour palette.

Stacked Bar Chart

Similarly to the heatmap, users can sort and order the features within the stacked bar charts in multiple ways, aggregate the results by cohort to look for trends in the data, manipulate the number of features shown, choose the level of taxonomy/hierarchy of the data and even randomize the colours for ease of interpretation.

Abundance Distribution

One of the more popular analyses in the Cosmos-Hub platform, the abundance distribution tool uses the Wilcoxon Rank-Sum test to reveal the statistical difference of any organism or functional gene/pathway between your cohorts in an interactive boxplot.

Principle Component Analysis (PCA)

Feature similarity across samples and cohorts are revealed in both 2D and 3D format

Alpha Diversity

Users have the ability to select between Chao1, Simpson or Shannon Diversity box plots and statistical significance testing using the Wilcoxon Rank-Sum test.

Beta Diversity

Available in both 2D and 3D format, PCoA plots for Jaccard and Bray-Curtis diversity are generated alongside PERMANOVA for significance testing.

LEfSe

Published by the Huttenhower lab at Harvard University in 2010, LEfSe (Linear discriminant analysis Effect Size) determines the features (taxa or functions) most likely to explain differences between cohorts when compared in a pairwise manner.

Multivariate Statistics

The Cosmos-Hub platform provides methods for multivariate statistics that enable users to conduct more personalized and sophisticated comparative analyses. With multivariate statistics, users can incorporate continuous metadata variables, control for confounders, run longitudinal analysis, select different models and transformations or even customize cutoffs for abundance, prevalence or significance.

Users have the flexibility to switch between different databases and metrics that have been run with a click of the button, easing the interpretation of the data. The outputs and underlying data exportable in a number of different formats including CSV data or PNG/SVG/PDF image formats.

Finally, when it’s time to publish, you can access each comparative analysis you’ve created in the main study dashboard where you will be able to access the methods and which parameters were used so that you or a peer can always reproduce the results.

Heatmap

A color-coded matrix that visually summarizes the strength and direction of associations between multiple metadata variables and microbial features, making it easy to spot overall patterns and clusters of significant relationships.

Box and Scatter Plots

Plots that display the distribution of microbial feature abundances or prevalences across different metadata categories (box plots) or along continuous variables (scatter plots), helping users visualize individual associations in detail.

Volcano Plots

A scatter plot that displays each feature’s effect size (fold change) versus its statistical significance, enabling quick identification of features with both strong effects and high significance.

Forest Plots

A plot showing effect sizes and confidence intervals for significant associations, allowing easy comparison of the magnitude and reliability of effects across multiple features and metadata contrasts.

Standardized & Reproducible

All comparative analysis pipelines are version-controlled to ensure standardization across your studies and between your team.

No-Code Platform

Run any statistical analysis tool without coding skills. All pipelines have been made available to run in a series of simple steps.

Interactive Charts & Visualizations

Cosmos-Hub makes interpretation of microbiome data easier by allowing users to manipulate figures in real time to reveal insights and communicate results more easily to the scientific community.

Convenience

All results are generated in parallel and aggregated in the study dashboard where users can toggle between outputs and analyses easily.

No Specialized Infrastructure Required

Save costs on servers and specialized computers by leveraging Cosmos-Hub’s cloud infrastructure. All you need is an internet connection and Google Chrome. That’s it!

Speed to results

Comparative analysis workflows and statistics are generated in just a few minutes, not hours, allowing users to complete their studies and publish in as much as half the time.

Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

Statistics Toolbox

Unlock the Microbiome Using the Cosmos-Hub Statistics Toolbox

How does Cosmos-Hub Support Virtual Cohort Analysis?

Single Study Analysis