Skip to content

Latest commit

 

History

History
14 lines (9 loc) · 817 Bytes

datasets.md

File metadata and controls

14 lines (9 loc) · 817 Bytes

Datasets

This work used three different datasets. One for training the models and two to validate the prediction models.

Dataset for training

We used American Gut Project (AGP) to train the model. Dataset was downloaded from AGP repository.

Datasets for validation

Two independent and external cohorts were used to validate our models.

  1. Chron´s subset from Gervers et al.
  2. Ulcerative colitis from Morgan et al.

Both datasets were retrieved from MLRepo repository created by KnightsLab. After clone MLRepo, you must to copy gevers and morgan folders into extdata folder.