Skip to content

load_example_data

source 

.load_example_data(
   omicstype = 'metabolomics', processed = True
)

Loads example datasets

Args

  • omicstype (str) : type of omics for example data. Available options are "metabolomics" or "transcriptomics". Metabolomics data are from Su et al 2020 https://doi.org/10.1016/j.cell.2020.10.037. Transcriptomics data - TO BE IMPLEMENTED
  • processed (bool) : Load processed (normalised, scaled) or raw data

Returns

pre-processed omics data matrix consisting of m samples and n entities (metabolites/genes) in the form of a pandas DataFrame. Contains one of more metadata columns at the end.

t_tests

source 

.t_tests(
   matrix, classes, multiple_correction_method, testtype = 'ttest'
)

Performs two-sample independent t-tests

Args

  • matrix (pd.DataFrame) : processed sample-by-compound metabolomics dataframe
  • classes (pd.Series) : pandas series containing phenotype metadata (e.g. 'COVID', 'NON-COVID')
  • multiple_correction_method (str) : see https://www.statsmodels.org/dev/generated/statsmodels.stats.multitest.multipletests.html for options
  • testtype (str) : Default is t-test, "mwu" also available to implement the Mann Whitney U test

Returns

pd.DataFrame containing p-values and corrected p-values for each metabolite

pathwaydf_to_dict

source 

.pathwaydf_to_dict(
   df
)

Converts pathway dataframe to dictionary, with pathway IDs as keys and metabolite lists as values

Args

  • df (pd.DataFrame) : Pandas DataFrame containing pathways

Returns

python dict pathway representation