/empress-analyses

Primary LanguageJupyter NotebookBSD 2-Clause "Simplified" LicenseBSD-2-Clause

empress-analyses

This document outlines the steps needed to recreate figure 1A, 1B and 2A-C. To generate the artifacts in these figures, we recommend installing Empress, cloning this repository, and following the steps in the Jupyter notebooks (see the notebooks folder).

For visualizing the QZV files we recommend using Safari or Firefox, since Google Chrome is known to show degraded performance when loading large files in certain cases (for example Figure 1A-B). See here for details on this.

Figure 1A

Link to QZV of EMPire plot.

  1. Change Layout to Circular
  2. Color tips by phylum assignment
    1. Feature Metadata Coloring color by phylum_assignment
    2. Set Color Map to Dark
    3. Click Update
  3. EMPO 1 barplot
    1. Check Draw barplots? in Barplots menu
    2. Click Sample Metadata
    3. Show sample info for empo_1
    4. Set Color Map to Yellow-Orange-Red
    5. Set Length to 150
    6. Click Update
  4. Search for specific feature
    1. Copy TACGTAGGTCCCGAGCGTTGTCCGGATTTATTGGGCGTAAAGCGAGCGCAGACGGTTACTTAAGCAGGATGTGAAATCCCCGGGCTCAAC into search bar
    2. Click Search

After configuring the tree, you can now configure the ordination panel on the right:

  1. Go to the Color tab on the right and select empo_1 as the Color Category (in the first dropdown menu).
  2. In the second dropdown menu in the Color tab, select Yellow-Orange-Red.

Figure 1B

Link to QZV of EMPress plot.

  1. Change Layout to Circular
  2. Phylum assignment barplot
    1. Check Draw barplots? in Barplots menu
    2. Click Feature Metadata
    3. Color by phylum_assignment
    4. Set Color Map to Dark
    5. Set length to 150
  3. Mean pH barplot
    1. Click Add another layer
    2. Click Feature Metadata
    3. Color by arithmetic_mean
    4. Set Color Map to Yellow-Green-Blue
    5. Check Continuous values?
    6. Set length to 150
  4. Add space between barplots
    1. Check Add a border around barplot layers?
    2. Set Length to 20
    3. Click Update

Figure 2A

Link to QZV of EMPress plot.

  1. In the Layout tab
    1. Change Layout to Circular
    2. Change Sort clades based on tip count? to No sorting
  2. In the Feature Metadata Coloring tab
    1. Select custom_level for Color by...
    2. Set Color Map to Spectral
    3. Check Collapse Clades
  3. In the Barplots tab
    1. Check Draw barplots?
    2. Click Feature Metadata
    3. Color by covid_hc_sig
    4. Set Color Map to Viridis
    5. Click Add another layer
    6. Color by covid_pn_sig
    7. Set Color Map to Viridis
    8. Check Add a border around barplot layers?
    9. Change Border Color to black
    10. Set border Length to 5
    11. Click the Update Button
    12. Note: No significant difference = 0, Lower in COVID-19 samples = -1, Higher in COVID-19 samples = 1
  4. Double click on the Carbon-carbon lyases clade (the larger red clade)
  5. In Search by node name..., search for 4.1.1.20

Figure 2B

Link to QZV of the EMPress plot.

To reproduce the figure within this QZV:

  1. Change Layout to Circular
  2. Color tips by their supperclass annotation:
    1. Use Feature Metadata Coloring to color by superclass.
    2. Set the Extra Line Width to 15, hit enter.
    3. Click Update.
  3. In the Barplots tab, click the checkbox marked Draw barplots?.
  4. Add a barplot for the common meal type annotation. Within the Layer 1 section:
    1. Click on Sample Metadata.
    2. Click on Show sample info for ... and select common_meal_type.
    3. Set the Color Map to Accent.
    4. Set Length to 700.
  5. Click Update.

See the notebook for details on how we merged the data, generated an EMPress visualization, and configured the visualization.

Figure 2C (and Supplemental Figure 1)

Link to QZV of the EMPress plot.

To reproduce the figure within this QZV:

  1. Change Layout to Circular
  2. Color tips by their phylum-level taxonomic classifications:
    1. Use Feature Metadata Coloring to color by Level 2.
    2. Set Extra Line Width to 15.
    3. Click Update.
  3. In the Barplots tab, click the checkbox marked Draw barplots?.
  4. Add a barplot for the Songbird differentials. Within the Layer 1 section:
    1. Click on Color by... and select C(brushing_event)[T.before].
    2. Set the Color Map to Red-Blue.
    3. Check the Continuous values? checkbox.
    4. Set Default length to 500.
  5. Click Update.
  6. Add a barplot for the ALDEx2 effect sizes. Click Add another layer. Within the Layer 2 section:
    1. Click on Color by... and select sfit.effect.
    2. Set the Color Map to Red-Blue.
    3. Check the Continuous values? checkbox.
    4. Set Default length to 500.
  7. Click Update.
  8. Add a barplot for the ANCOM W-statistics. Click Add another layer. Within the Layer 3 section:
    1. Click on Color by... and select W_stat.
    2. Set the Color Map to Purples.
    3. Check the Continuous values? checkbox.
    4. Set Default length to 500.
  9. Click Update.
  10. Add a border around each barplot layer:
    1. Click the Add a border around barplot layers? checkbox.
    2. Set Length to 100.
  11. Click Update.

See the notebook for details on how we merged the data, generated an EMPress visualization, and configured the visualization.