Course development ideas

[tavareshugo]

Below is an outline for a 4-to-5 day course on bacterial genomics. However, the course could run as a 2-to-3 day course if focusing only on TB.

---

Prerequisites

• Unix
• R (preferable)

---

Day 1

• Intro slides covering:
  • know your bug
  • diagram of different paths of analysis (which shows up through the workshop to keep participants oriented in the great scheme of things)
  • brief intro to Nextflow (nf-core)
  • Mapping-based assembly
• Practical focuses on nf-core/bactmap workflow
• TB-focused examples
• At the end of the day participants should have the output of running bactmap
  • explore multiqc output

Day 2

Continuing with TB data

• Phylogenetics
  • recap slides about phylogenetic inference
  • practical using IQtree2 (alignments come from bactmap)
• TB profiler (assigning lineages, mutations that confer AMR)
• Microreact for displaying trees and metadata
• Transmission dynamics
  • networks of transmission

Day 3

Switching to reference-independent assembly

• de-novo assembly, annotation
  • nf-core/bacass workflow (?)
  • Staphylococcus (not as much recombination) or Streptococcus (more recombination)
• Panaroo for pan-genome analysis
  • generate core genome alignments (genes that exist in high proportion of the genomes)
  • accessory genome (no analysis will be done on this, although potentially there could be stuff to explore there)
  • highlight that this tool takes a long time to run (e.g. need long queue on HPC or run in chunks and merge)
• Phylogenetic tree built on the core genome alignment
  • practical with IQtree (alignments come from panaroo)
• Downstream analysis with Pathogenwatch (takes the FASTA as input)
• Tree visualisation with Microreact (using the metadata from Pathogenwatch)

Day 4

Continuing with the recombining bacteria from day 3

• Clustering on phylogenetic tree
  • fastbaps
• bactmap on each cluster
• recombination removal
• iqtree
• transmission analysis
  • building network of transmission (nodes are patients, links are putative tranmission path weighted by pairwise distances)
• Integration with epi data?
  • supplement networks of transmission with epi data