Practical 1b: Multilocus Phylogeny-Based Placement
This tutorial provides step-by-step instructions for phylogenetic placement and analysis using T-BAS and DeCIFR.
Practical 1b Workflow.
- In this tutorial, you will use multiple loci to place unknown fungal sequences on a reference phylogeny.
- Select the Fungi v3 reference tree
- Download the example multilocus files
- Upload the six FASTA files
- Upload the metadata
- Configure and run EPA-ng phylogenetic placement
- Inspect multilocus placements and support
Goal: evaluate whether multiple loci support consistent placement of the unknown sequences.
Screenshot guide
Learning Outcomes.
By the end of this practical, you should be able to:
- Set up a multilocus phylogenetic placement analysis in T-BAS.
- Recognize how loci are matched to the reference tree.
- Use EPA-ng to place unknown sequences on an existing phylogeny.
- Inspect placement support in the tree viewer.
- Inspect sequence evidence across loci to identify consistent or potentially conflicting signals.
Screenshot guide
Why Use Multiple Loci for Phylogenetic Placement?
Single-locus placement
A single barcode can provide useful taxonomic information, but its resolution may be limited for closely related taxa.
Multilocus placement
Multiple loci provide independent phylogenetic evidence that can be compared for the same unknown sample.
Ask whether the loci tell the same evolutionary story.
Consistent placements strengthen an identification; conflicting or weakly supported placements signal uncertainty that should be investigated.
Screenshot guide
Part 1 — Set Up and Run the Multilocus Placement Analysis
Place unknown sequences on a reference phylogeny and evaluate placement confidence and support.
Step 1. Select the Reference Tree.
What you will do
- Open the T-BAS phylogenetic placement workflow.
- Click to select a reference tree.
- Choose Fungi v3 for the example analysis.
Why this matters
- The reference tree determines which taxa and loci are available for phylogenetic placement. Your query sequences are interpreted in the context of this existing phylogeny.
- For this tutorial: select Fungi v3.
Screenshot guide
Open the T-BAS Phylogenetic Placement Workflow.
- Click here to select reference tree
Screenshot guide
Select the Fungi v3 Reference Tree.
Screenshot guide
Step 2. Download the Example Files.
Download the example files
- Six FASTA files containing the unknown sequences
- The accompanying metadata file
- The reference paper provided with the example
- Optional ZIP archive containing all example files
Before continuing
- Keep the six locus files separate. In the next step, they will be uploaded together as separate unknown-query files, with ITS placed first.
- Also keep the metadata file available; it will be uploaded separately.
- These files provide the complete example dataset used throughout Practical 1b.
Screenshot guide
Download example files.
- Click on examples
Screenshot guide
Download the Example Files.
Download:
- six FASTA files
- metadata
- reference paper
- Optional: ZIP archive containing all files
Screenshot guide
Step 3. Upload the Multilocus Query Sequences.
What you will do
- Drag the six FASTA files into the unknown query box.
- Place the ITS file at the top of the list.
- Confirm that all six sequence files have been added before continuing.
Why file order matters
- For multilocus datasets, ITS must be the first FASTA file for T-BAS to run the ITS-specific UNITE BLAST analysis and generate the UNITE report.
Checkpoint: six FASTA files should be visible in the unknown query input.
Screenshot guide
Drag the Six FASTA Files into the Unknown Query Box.
- Confirm ITS is at the top
Screenshot guide
Step 4. Upload the Metadata.
What you will do
- Drag the provided metadata file into the metadata input box.
- Confirm that the metadata file is associated with this placement run.
Why metadata matter
- Metadata provide sample labels and other information that can be used to visualize and interpret the phylogenetic placements.
- Sequence data and metadata are uploaded separately and combined during the analysis.
Screenshot guide
Drag the Metadata File into the Box.
- Upload metadata.csv
Screenshot guide
Step 5. Configure the Placement Analysis.
Select the analysis options
- Choose EPA-ng (Evolutionary Placement Algorithm – Next Generation) for phylogenetic placement.
- Select “ITS locus is included (first FASTA) – generate UNITE report.” to BLAST the ITS sequences against UNITE.
- Provide a descriptive label for the run.
Placement strategy
- EPA-ng places each query sequence onto the existing reference phylogeny without rebuilding the tree.
- Placements can then be examined in the T-BAS tree viewer.
Screenshot guide
Select the Option to Generate the UNITE Report.
Screenshot guide
Select EPA-ng for Phylogenetic Placement.
Screenshot guide
Provide a Descriptive Label for the Run.
- Enter a run label
Screenshot guide
Confirm Automatic Locus Selection.
Check the locus assignments
- T-BAS automatically selects loci when uploaded filenames match locus names in the reference tree.
- Verify that each FASTA file is assigned to the expected locus before starting the run.
Why check this?
- Correct locus matching ensures that each query sequence is compared with the corresponding locus represented in the multilocus reference dataset.
- Do not assume automatic selection is correct—verify the locus assignments before proceeding.
Screenshot guide
Confirm That Each Locus Was Selected Correctly.
- Verify the locus assignments
Screenshot guide
Disable ITSx for This Multilocus Example.
What you will do
- Cancel or disable the ITSx step for this multilocus example.
- Leave the remaining placement settings as specified in the tutorial.
Tutorial focus
- In Practical 1a, ITSx was used to confirm that ITS sequences did not contain large SSU or LSU overhangs that could interfere with alignment to the ITS reference sequences. The ITS sequences provided for this tutorial are already suitable for alignment, so ITSx can be skipped.
- The goal here is to compare phylogenetic evidence across loci.
Screenshot guide
Click Cancel to Skip ITSx.
Screenshot guide
Run the Analysis and Monitor Progress.
Start the placement run
- Submit the configured EPA-ng analysis.
- Use the run status page to monitor progress.
- Wait until the analysis is complete before opening the placement tree (typically takes about 10 min to complete).
Next
- After the run finishes, you will switch to the rectangular tree layout and locate the example unknowns to evaluate their phylogenetic placements.
- Next section: Explore and interpret the multilocus placements.
Screenshot guide
Monitor the Placement Run.
Screenshot guide
Part 2 — Explore Multilocus Phylogenetic Placements
Place unknown sequences on a reference phylogeny and evaluate placement confidence and support.
Step 6. Switch to the Rectangular Tree Layout.
What you will do
- Open the completed phylogenetic placement result.
- Switch the tree viewer to the rectangular layout.
Why use this view?
- The rectangular layout makes it easier to follow branches, read taxon labels, and see where unknown sequences fall relative to named reference taxa.
- Use the tree as a map: first locate the unknown, then examine its nearest reference taxa.
Screenshot guide
Switch view to rectangular layout.
Screenshot guide
Locate the "Bipolaris" Unknown.
Search the placement tree
- Enter “Bipolaris” in the search box.
- Locate the highlighted unknown sequence on the reference tree.
- Identify the closest named reference taxa.
As you inspect the placement, ask:
- Is the unknown nested within or adjacent to a named lineage?
- What taxonomic assignment would you make from its position?
- How confident are you based on tree position alone?
- Record your initial taxonomic assignment before examining LWR support.
Screenshot guide
Search “Bipolaris” to locate the “test” unknown.
Screenshot guide
Locate the "Aspergillus" Unknown.
Repeat the search
- Enter “Aspergillus” in the search box.
- Locate the unknown sequence and inspect its surrounding clade.
- Compare its placement with the Bipolaris example.
Compare the examples
- Where does each unknown fall relative to named reference taxa?
- Does one appear easier to identify than the other?
- What taxonomic assignment would you make from tree position alone?
- Now test these initial interpretations using EPA-ng and gappa LWR support.
Screenshot guide
Enter ”Aspergillus” in the search box to locate the “test” unknown.
Screenshot guide
Part 3 — Compare Placement Support
Integrate phylogenetic placement with multiple reference databases to improve taxonomic resolution and evaluate assignment support.
Display and Interpret EPA-ng and gappa LWR Support.
Adjust the tree display
- Uncheck branch lengths.
- Turn on EPA-ng LWR Support.
- Turn on gappa LWR Support.
- Re-examine the placement of each unknown.
Interpretation
- EPA-ng LWR: How strongly is the query supported at its placement?
- gappa LWR: How strongly does the placement support the taxonomic assignment?
Interpret the two values together.
- Strong placement support does not necessarily mean a confident taxonomic assignment.
Screenshot guide
Select EPA-ng and gappa LWR support, then scroll to the legends.
Screenshot guide
Re-examine “Aspergillus” using EPA-ng and gappa LWR.
Screenshot guide
View EPA-ng and gappa LWR Values in the Taxonomy Resolver.
- Run the Taxonomy Resolver using the run accession to view the numerical EPA-ng and gappa LWR values for each query.
- The tree provides a visual summary; the Taxonomy Resolver provides the numerical EPA-ng and gappa LWR values.
Screenshot guide
In-Class Discussion: Compare the Placements.
- Use the Bipolaris and Aspergillus examples to discuss how multilocus evidence should be interpreted.
- Where was each unknown placed relative to named reference taxa?
- What do the EPA-ng and gappa LWR values tell you about placement and taxonomic confidence?
- Do the individual loci show sequence patterns consistent with the inferred placement?
- Would you assign the unknown to a species, a broader lineage, or leave it unresolved?
Take-home message: Taxonomic resolution should reflect both placement support and assignment support.
Screenshot guide
