Skip to content

Methylation Task: Lineage Assignment

Warning

Use the methylation Codespace for this task. It's the same one from the methylation practicala as it already contains the data you need.

Introduction

In this task you'll combine two datasets, the Nanopore barcodes and the PacBio strains, into a single phylogeny, and use it to work out the lineages of the PacBio samples.

The Nanopore samples already have lineage assignments (from tb-profiler). The PacBio samples don't. By placing both sets of sequences on one tree, you can infer each PacBio strain's lineage from the Nanopore samples it clusters with.

Objective

Assign a lineage to each PacBio sample by building a combined phylogeny and reading lineages across from the Nanopore samples.

Starting point

cd data/methylation/tree/

Have a look at what's already there before you start (ls).

Steps

You'll need to work out the exact commands yourself, that's the task, but here's the route:

  1. Combine the sequences. You have separate FASTA files for the Nanopore and PacBio samples. Merge them into one multi-FASTA.
  2. Align. The combined sequences need to be aligned together with mafft before you can build a tree.
  3. Build the tree. Infer a phylogeny from the alignment (the same tree-building tool you used in the phylogenetics practical works here).
  4. Annotate in iTOL. Upload the tree, then bring in the annotations:
    • the lineage of each sample, and
    • which platform (Nanopore or PacBio) each sample came from. You have an annotation file for each dataset, you'll need to combine the two annotation files into one so every tip on the tree is labelled.
  5. Read off the answer. Find where each PacBio sample sits and assign it the lineage of the Nanopore clade it falls within.

Hints

  • The tree-building and mafft steps mirror what you did in the phylogenetics practical, reuse those commands.
  • For the annotations: both files are the same format (tip label + value). Combining them is just a matter of getting all tips into one file with consistent columns, watch that the tip labels match the names in your tree exactly, or iTOL won't attach them.
  • A PacBio sample sitting inside a Nanopore lineage clade takes that lineage. One sitting on its own long branch, between clades, is ambiguous, note it rather than forcing an assignment.