Protistan predation pressure at hydrothermal vents

Associated code for data analysis, amplicon sequence processing, and figure generation for manuscript entitled Protistan grazing impacts microbial communities and carbon cycling at deep-sea hydrothermal vents.

Hu, S.K., Herrera, E., Smith, A., Pachiadaki, M.G., Edgcomb, V.P., Sylva, S.P., Chan, E.W., Seewald, J.S., German, C.R., & Huber, J.A. (In Press) Protistan grazing impacts microbial communities and carbon cycling at deep-sea hydrothermal vents.

• PNAS
• Preprint available

All code and analyses presented here.

Description of contents

1. Study background 🌊🌋

Microbial eukaryotes (or protists) in marine ecosystems are a link between microbial primary producers and all higher trophic levels. The rate at which heterotrophic protistan grazers consume microbial prey and recycle organic matter is an important component of marine microbial food webs and carbon cycling. At deep-sea hydrothermal vents, chemosynthetic bacteria and archaea form the basis of a food web in the absence of sunlight, but the role of protistan grazers in these highly productive ecosystems is largely unexplored. Code presented here is for the analysis of grazing experiments, where hydrothermal vent fluid was used in grazing incubations, and amplicon sequences to characterize the protistan and prokaryotic communities associated with the hydrothermal vent environment.

2. Working R environment 💻🧶

Using R version 3.6.1 with RMarkdown.

3. Grazing experiment 🧪🔬

Description of code that imports raw cell count information (derived from microscopy counts) and processes this information to determine FLP per eukaryotic cell and downstream estimates of grazing impact.

• Import results from cell counts (cells / ml)
• Quality check input data and reformat
• Determine microscopy count error rate
• Determine which time points have significant differences in FLP over time
• Calculate grazing effect, rate, and prokaryote turnover percentage
• Estimate percent turnover with respect to carbon
• Generate Figure 1

4. Process 18S rRNA gene amplicons 🧬🧼

Imports raw Amplicon Sequence Variant (ASV) count files, conducted 'decontam' to remove potential contaminate ASVs, and quality checks taxonomy assignment.

• Import and reformat ASV table
• Use decontam to remove contaminants
• Reformat for downstream analyses

5. Characterize protist diversity 🧰📊

Analysis of protist community structure and diversity by averaging across replicate samples, summing to individual taxonomic groups for visualization purposes, and ordination analysis to explore sample-to-sample differences.

• Curation of assigned taxonomies
• Reformat counts to determine average counts across replicates
• Perform PCA analysis
• Generate Figure 2

6. Classify protist distribution 🌊🍸

18S rRNA gene derived ASVs were classified based on their distribution among samples. ASVs found throughout the entire hydrothermal vent system, including the background seawater environment were classified as 'cosmopolitan'. 'Resident' ASVs were those found only within the hydrothermal vent fluid.

• Categorize ASVs based on distribution
• Plot ASV distribution by total ASVs and total sequences

7. Protistan distribution and diversity across samples 🗂️🌊

Code to format taxonomic classification to highlight approximately family or class level, including ASV richness (total ASVs).

• Re-curate taxonomy
• Determine ASV richness
• Perform CLR transformation
• Generate tile plot (Figure 3)

8. Process 16S rRNA gene amplicons 🦠🧬

Similar to 18S rRNA gene tag-sequencing pipeline. Import raw ASV count tables and characterize prokaryote community.

• Curate 16S taxonomy
• Generate barplot
• Perform CLR transformation and PCA

9. Network analysis 🕸️

Import 16S and 18S datasets as phyloseq objects and subsample to select more abundant ASVs from in situ samples. Explore and run SPIEC-EASI

• Run multi-domain SPIEC-EASI analysis (used HPC)
• Export results as phyloseq object to import back into R locally
• Import into R
• Parse significant pairs of 18S-16S ASVs
• Generate alluvial donut plots to show relationship between putative predator and prey (Figure 4).

10. Regression analysis 📈

Explore relationship between geochemistry information from each vent site and results from grazing analysis.

• Import and reformat grazing results and environmental parameters
• Perform linear regression, pull out slope, r2, and intercept
• Generate plot to explore relationships

Graphical comic abstract

Last updated

Feb 4, 2021

Notes on RMarkdown compilation

Instructions adapted from here.
Steps: Ahead of time, make sure the repo is started by adding R as an .ignore option.

1. From cloned repo, open RStudio project with all contents, update .Rmd file
2. Save and use Knitr to compile .html and update index files stored in /docs
3. These /docs files were made possible by including this header line in the RMarkdown file:

knit: (function(input_file, encoding) {
  out_dir <- 'docs';
  rmarkdown::render(input_file,
 encoding=encoding,
 output_file=file.path(dirname(input_file), out_dir, 'index.html'))})

4. Once changes have been saved and Knit complete, commit and comment all changes to git
5. push changes
6. Under settings, set GitHub Pages Source to main branch and /docs