zaneveld / gcmp_global_disease Goto Github PK

Is the latter just the former imported into QIIME2? Or are there other quality control changes?

To generate a new trait table of diseases x species, we need to map the disease data from the FRRP, HICORDIS and Lamb lab databases onto the tree names for the Huang-Roy tree, using the new species name mappings provided by #26 in metadata.

Hannah is working on a Generalized Diversity Modeling analysis based on the code from fungal endophyte modeling in Darcy et al., 2020

Breaking apart the giant metadata sheet will allow us to more easily update the coral species' traits reliably across the table.

In the SILVA insertion tree, we explicitly state the reference phylogeny. Let's update the Greengenes insertion tree to have GG or Greengenes in the title .

...this is mostly the fault of the adiv and bdiv script. It's outputs should be saved in a subfolder similar to the effects_of_rarefaction ipynb

Ideally we want a method that can incorporate fancier linear models (not just univariate analysis). Candidates: ANCOM? ALDEx? Songbird? Birdman?

To match up the Lamb et al genus level data to our tree:

1. Reconstruct microbiome richness on the tree
2. Iterate over all genus names (obtained by splitting the tree tip name and taking 1st element)
3. For each genus, find the LCA of all tips with a genus name matching that genus
4. The ASR estimate for that internal node for the genus becomes the 'richness of the genus'
5. Probably we should prune the tree back to the genus level tips, more of which should now have both disease (Lamb et al) and microbiome data.
6. For HICORDIS, FRRP, and Page and Willis, should we do the same thing with the disease data to get genus level estimates?

This should be a .csv mapping between the coral species names from the HICORDIS, FRRP, and Lamb lab disease datasets to the Huang Roy tree names. The purpose is to allow a unified analysis between microbiome, phylogenetic and disease datasets.

The format could be:

Huang_Roy_tree_name\tFamily\tGenus\tSpecies\tDisease_Project\tDisease_Name\tPrevalence\tCount_Diseased\tCount_Not_Diseased

M_ft.qza
T_ft.qza
S_ft.qza

We need to write a Jupyter notebook that uses SEPP insertion (through the QIIME2 python API) to generate an updated tree, then calculate phylogenetic alpha and beta diversity metrics

Strategy:

Set the comment.chars parameter on read.table to ''
Rename the X.SampleID column after read.table converts #SampleID to X.SampleID

OR revise metadata to remove the # if QIIME2 accepts this

Try running the coral disease processing steps from top to bottom, revising so that we can easily replicate the whole workflow.

-- Check that all outputs are in output
-- Remove unnecessary folders (git rm)