1 of 8

Applied Bioinformatics 2025�Week 2 Session 1�Data visualization with mixOmics

Natalie Turner, PhD

Postdoctoral Fellow – Yates Lab

Department of Molecular Medicine

naturner@scripps.edu

2 of 8

Data visualization with mixOmics

Principal component analysis (PCA)
Sparse (or supervised) Partial Least Squares Discriminant analysis (sPLS-DA)
Clustered Image Map (CIM)
Correlation Circle Plot
Many others not covered here

3 of 8

Formatting

mixOmics requires specific formatting!
It also requires pre-processed data as input (as per week 1), including normalization (we’ll cover this in the next session)
Today, we will get acquainted with mixOmics and its functionality

4 of 8

PCA

Think of it like, ‘getting the lay of the land’.

A Principal Component Analysis (PCA) plot is like a map for your data. It takes complex data and reduces its dimensions, making it easier to see patterns and relationships.

5 of 8

sPLS-DA

Key Features: It shows which features (or variables) are most important for differentiating between groups.

A Partial Least Squares Discriminant Analysis (PLS-DA) plot is like a roadmap that helps you see how different groups of data points are related to each other.

6 of 8

PCA vs (s)PLS-DA

PCA

Uncovers the overall structure and patterns in data without prior group information

(s)PLS-DA

Distinguishes between specific groups and highlights differences based on predefined classes.

7 of 8

Components

“Components are linear combinations of features.

Loadings represent the weights (or coefficients) assigned to each of the features to determine their contribution to a given component.”

https://mixomics.org/faq/glossary/#:~:text=Loadings,contribution%20to%20a%20given%20component.

8 of 8

Go to your R Notebook

Follow the instructions to work on a breast cancer dataset from The Cancer Genome Atlas network, ‘breast.TCGA’
Take note of the data formatting – how is the dataset structured in R? What information does it contain?