gavinsimpson / physalia-multivariate

Multivariate analysis of biological data

Multivariate data analysis with R and vegan

Physalia-Courses

https://www.physalia-courses.org/

Gavin Simpson

27 November – 1st December, 2023

Overview

The R statistical language has enjoyed wide and rapid adoption by many ecologists, and is used across many ecological subdisciplines for statistical analyses and the production of publication-quality figures. For community ecologists using R, one of the most-used, and useful, add-on packages is vegan, which provides a wide range of functionality covering inter alia ordination, diversity analysis, and ecological simulation. This workshop will offer participants a practical introduction to some of the most useful functions available within vegan. We will focus on the use of ordination methods and on the use of restricted permutations to test a range of experimental designs.

We will focus in particular on when and how to use multivariate methods including unconstrained and constrained ordination (CCA, RDA, Constrained PCoA), as well as between-group tests such as PERMANOVA. We will cover concepts such as design- and model-based permutations and the exchangeability of samples in tests. We will also discuss the use of vegan to go beyond simply fitting a constrained ordination model, to diagnostics, plotting, etc.

Slides

• Monday
  • PDF
• Tuesday
  • PDF
• Wednesday
  • PDF
  • Ponds example
  • Spring meadows example
• Thursday
  • PDF
  • Randomised Block
  • Permutation test examples
• Friday
  • PDF
  • PERMANOVA
  • dbRDA
  • PRC
  • Co-CA
  • Variation partitioning
  • Linear discriminant analysis

Target audience and assumed background

This course is suitable for PhD students (including senior thesis-based masters students) and researchers working with multivariate data sets in biology (inter alia ecology, animal science agriculture, microbial ecology/microbiology), with limited statistical knowledge but a willingness to learn more.

Participants should be familiar with RStudio and have some fluency in programming R code, including being able to import, manipulate (e.g. modify variables) and visualise data. There will be a mix of lectures, and hands-on practical exercises throughout the course.

Learning outcomes

1. Have a good introductory understanding of the main approaches used in the analysis of multivariate data sets,
2. Be able to choose an appropriate method to use to analyse a data set
3. Understand how to use restricted permutation tests with constrained ordination methods to test the effects of predictor variables or experimental treatments,
4. Be able to use the R statistical software to analyse multivariate data

Pre-course preparation

R

Please be sure to have at least version 4.2 — and preferably version 4.3 — of R installed. Note that R and RStudio are two different things: it is not sufficient to just update RStudio, you also need to update R by installing new versions as they are released.

To download R go to the CRAN Download page and follow the links to download R for your operating system:

• Windows
• MacOS X
• Linux

To check what version of R you have installed, you can run

version

in R and look at the version.string entry (or the major and minor entries).

RStudio

We will use RStudio as the environment for interacting with R. You are free to use your preferred IDE or R interface if you wish. To install RStudio, go to https://posit.co/download/rstudio-desktop/ and follow the instructions.

Required R Packages

We will make use of several R packages that you'll need to have installed. Prior to the start of the course, please run the following code to update your installed packages and then install the required packages:

# how many CPU cores are available?
cores <- parallel::detectCores()
# update any installed R packages
update.packages(ask = FALSE, checkBuilt = TRUE, Ncpus = cores - 1)

# packages to install
pkgs <- c("vegan", "tidyverse", "cocorresp", "permute", "here", "tibble",
          "readxl", "janitor", "stringr", "ggrepel")

# install those packages
install.packages(pkgs, Ncpus = cores - 1)

Please also install my in-development ggvegan package as it will be useful to know how to draw ordination diagrams using the ggplot2 package. As ggvegan is not yet on CRAN, we need to install from my R Universe:

# Enable repository from gavinsimpson
options(repos = c(
  gavinsimpson = 'https://gavinsimpson.r-universe.dev',
  CRAN = 'https://cloud.r-project.org'))
# Download and install ggvegan in R
install.packages("ggvegan")

Setting the options like this only affects the current R session, which is all we need to install ggvegan.

Ideally, we will also make use of the DESeq2 package, which has to be installed from the Bioconductor project:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("DESeq2")

But don't worry if you can't get this installed, we won't be using it for any of the longer computer examples, just to illustrate a couple of workflows for handling high-throughput data.