mimic-da 0.0.0

No project description provided

miMic (Mann-Whitney image microbiome)

This code is attached to the paper "miMic - a novel multi-layer statistical test for microbiome disease". miMic is a straightforward yet remarkably versatile and scalable approach for differential abundance analysis. miMic consists of three main steps:

(1) Data preprocessing and translation to a cladogram of means.

(2) An apriori nested ANOVA (or nested GLM for continuous labels to detect overall microbiome-label relations.

(3) A post hoc test along the cladogram trajectories.

How to apply miMic

miMic is available at this GitHub, PyPi, and in the following website.

miMic's GitHub

There is an example in example_use.py. You should follow the following steps:

1. Load the raw ASVs table in the following format: the first column is named "ID", each row represents a sample and each column represents an ASV. The last row contains the taxonomy information, named "taxonomy".

df = pd.read_csv("example_data/for_process.csv")

2. Load a tag table as CSV, such that the tag column is named "Tag".

tag = pd.read_csv("example_data/tag.csv",index_col=0)

3. Apply the MIPMLP with the defaulting parameters (see MIPMLP PyPi or MIPMLP website for more explanations).

   processed = MIPMLP.preprocess(df)
   

4. micro2matrix (translate microbiome into matrix according to iMic, and save the images in a prepared folder.

 folder = "example_data/2D_images"
  samba.micro2matrix(processed, folder, save=True)

Note for more information on SAMBA and for further distance calculations.

5. Apply the miMic test. One can choose the following hyperparameters:

   • eval (evaluation method) Choose one of "mann" for binary labels, "corr" for continuous labels, and "cat" for categorical labels.

   • sis (apply sister correction) Choose one of "bonferroni" (defaulting value) or "no".

   • correct_first (apply FDR correction to the starting taxonomy level) Choose one of True (defaulting value) or False.

   • mode (2 different formats of running) Choose one of "test" (defaulting value) or "plot". The "plot" mode should be applied only if the "test" mode is significant.

   • save (whether to save the corrs_df od the miMic test to computer) Choose one of True (defaulting value) or False.

   • tax (Starting taxonomy of the post hoc test) Choose one of None ((defaulting value for "test" mode) or one of 1, 2, 3, 4, 5, 6, 7. In the "plot" mode the tax is set automatically to the selected taxonomy of the "test" mode.

      taxonomy_selected = apply_mimic(folder, tag, eval="man")
      if not taxonomy_selected:
        apply_mimic(folder, tag, mode="plot", tax=taxonomy_selected, eval="man")
     

miMic's PyPi

1. Install the package

      pip install mimic