16S Data Analysis Part I

Elizabeth Costello, Ph.D. - Stanford University

Workshop materials can be found here:

https://elsherbini.github.io/durban-data-science-for-biology/

Practicalities

WiFi:

Network: KTB Free Wifi (no password needed)

Network AHRI Password: @hR1W1F1!17

Network CAPRISA-Corp Password: corp@caprisa17

Bathrooms are out the lobby to your left

Hello đź‘‹đź‘‹

My name is Elizabeth Costello (Liz)

  • I’m a Research Scientist in David Relman’s lab at Stanford University’s School of Medicine.

  • I enjoy microbial ecology, bioinformatics AND spending time outdoors.

Liz kayaking

Workflow check-in

Let’s review what we’ve done and what’s next

  • Done: Starting with the raw sequence data, we filtered, trimmed, inferred ASVs, removed chimeras, and assigned taxonomy.

  • Next up: Take the results of these steps (count table; taxonomy table) and do some initial and exploratory data analysis.

🎯 Goals for this module

  1. Understand what exploratory data analysis is

  2. Apply exploratory data analysis to 16S data

    1. Get to know the data
    2. Create a phyloseq object
    3. Explore (stacked) bar plots
    4. Explore alpha (within-sample) diversity
    5. Explore beta (between-sample) diversity using ordination

  3. Become familiar with the phyloseq package

Exploratory data analysis

A popular quote:

“Exploratory data analysis is an attitude, a flexibility, and a reliance on display, NOT a bundle of techniques” -John Tukey

🫣 What do the data look like? 🔬

Key considerations for molecular survey data

  • Uneven/incomplete sampling depth
  • Wideness/sparseness
  • Compositionality
  • Collinearities
  • Confounders
  • “Garbage in, garbage out”

The phyloseq package

McMurdie & Holmes (2013)

Alpha diversity (𝜶)

Fierer et al. (2012)
  • Within-sample diversity
  • Richness: The number of taxa or lineages within a given sample
  • Evenness: The relative abundance of taxa or lineages within a given sample
  • A metric that considers both richness & evenness (e.g., the Shannon diversity index)

Beta diversity (β)

Fierer et al. (2012)
  • Between-sample diversity
  • Taxonomic or phylogenetic difference in community composition between samples
  • Various measures of pairwise resemblance
  • Dimensionality reduction (e.g., ordination)
  • Identify primary sources of variation in data

Let’s analyze some data!

Go to RStudio, navigate to directory 16S_part1, and open Quarto document 16S_part1_code.qmd

back to module