[Networking] Advanced Qiime2 Integration

[mwang87]

1. Bring data with metadata
2. Explore qiime2 capability and define analyses that can be performed on certain classes of data
3. Produce qiime2 command lines to produce at least once what we would want to provide
4. Test and validate results on your data

[mwang87]

Next Steps

• [x] Find example data to analyze with SEED projects - @kcweldon20
• [x] Run through GNPS to create qza and manually create permanova @kcweldon20
• [x] Exploratory bi-plot analysis @kcweldon20 @alan-jarmusch
• [x] Prototype Parsing of Metadata file to determine columns to consider for Permanova @cmaceves
• [x] Determine what to give to user: qzv, qza, extracted table, naming? @fernie0341

[cmaceves]

@mwang87 in which directory should I put the metadata parser script?

[mwang87]

@cmaceves

https://github.com/CCMS-UCSD/GNPS_Workflows/tree/master/feature-based-molecular-networking/tools/feature-based-molecular-networking/scripts

[kcweldon20]

• [ ] Chrissy writes metadata script
• [x] Ming implements Biplot (Kelly provides code) test link
• [ ] Alan and Ming figure out how to put masses or annotations on the arrows - Punting on this for now
• [X] Kelly provides code for PERMANOVA
• [x] Fernie naming stuff (outputs)
• [X] Kelly provides GNPS job to work with (Sejal)

[kcweldon20]

Example Dataset: Seed Grant B0016 Disease-associated microbial species, for disease diagnosis and prevention of Type 1 diabetes in children ages 7 to 18. https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=c29599fbe6cf4425b5ab8fc10eb6985d

[kcweldon20]

PCoA Biplot Code

1) Calculate Beta Diversity

qiime diversity beta \
 --i-table categorical_ms2.qza \
 --p-metric canberra \
 --o-distance-matrix canberra_qiime2.qza

2) Compute PCoA

qiime diversity pcoa \
  --i-distance-matrix canberra_qiime2/distance_matrix.qza \
  --o-pcoa pcoa_canberra_qiime2.qza

3) Compute Relative Frequency Table for Biplot

qiime feature-table relative-frequency \
  --i-table categorical_ms2.qza \
  --o-relative-frequency-table relative_frequency.qza

4) Compute PCoA Biplot

qiime diversity pcoa-biplot \
  --i-pcoa pcoa_canberra_qiime2.qza \
  --i-features relative_frequency.qza \
  --o-biplot pcoa_biplot_canberra_qiime2.qza

5) Visualize PCoA Biplot

qiime emperor biplot \
  --i-biplot pcoa_biplot_canberra_qiime2.qza \
  --m-sample-metadata-file metadata.txt \
  --p-number-of-features 10 \
  --o-visualization pcoa_biplot_emperor_qiime2/visualization.qzv

[kcweldon20]

PERMANOVA Code (Per category) Using Beta Diversity Calculation

  qiime diversity beta-group-significance \
  --i-distance-matrix canberra_qiime2.qza \
  --m-metadata-file metadata.txt \
  --m-metadata-column COLUMN \
  --o-visualization Column_PERMANOVA.qzv

[fernie0341]

I would recommend that we change the GNPS output links as follows

Advanced Views - QIIME2 Views

View qiime2 PCoA Plot (changed Emperor to PCoA) Download qiime2 PCoA qzv (changed Emperor to PCoA) View qiime2 PCoA Biplot Download qiime2 PCoA Bilplot qzv View qiime2 Permanova (this option would be used if we visualize the permanova on GNPS) Download qiime2 Permanova qzv Download qiime2 Permanova tsv

[mwang87]

What are the filenames? What is the navigation structure? Which Permanova? We can generate it for a variety of columns.

[kcweldon20]

We need to discuss the distance matrix to use on the PCoA and PCoA biplot outputs from GNPS. Can also discuss the idea of multiple outputs.

[cmaceves]

I was going to use a Euclidean distance matrix for the permanova implementation.

[mwang87]

We have included the output in FBMN to directly link to the biplot in view.qiime2.org

[mwang87]

Separating this issue out https://github.com/CCMS-UCSD/GNPS_Workflows/issues/258 https://github.com/CCMS-UCSD/GNPS_Workflows/issues/259