/biomarkers

Algorithm for discovering combinatorial biomarkers

Primary LanguageJupyter Notebook

Combiom: an algorithm to search for combinatorial biomarkers

Implementation

The algorithm to search for combinatorial biomarkers is incorporated in Combiom. searching for combinatorial biomarkers.ipynb notebook and available as a separate combiom.py file. An example of data analysis is available in Combiom. Statistical analysis, cross validation and plotting.ipynb.

Usage

Combiom can be imported by:

import combiom as cb

Firstly, load some data and convert it to numpy arrays. Then call init_iterators() and search() functions:

# Initializing with 5 parameters
iters = cb.init_iterators(5)

# Searching for combinatorial biomarkers
results = cb.search(bio_marker_names, bio_marker_data, bio_target_names.size, bio_target_data, bio_target_names, iters)

Results

Resulting Pandas DataFrames are stored as pickle files. Excel files were created with Pandas and XlsxWriter. See Jupyter notebooks for details.