Human Diabetes Glycomics (ML Ready)

The Human Diabetes Glycomics (ML Ready) dataset contains abundance information on the plasma N-glycome of individuals participating in the FinRisk study and can be used to predict the risk of incident diabetes within a 10-year follow-up period. Baseline samples were analyzed from 37 individuals who developed type 2 diabetes within 10 years and 37 sex and age matched individuals who remained normoglycaemic. HILIC-UPLC chromatograms were separated into 46 peaks and were assigned to the most abundant glycan structures in each peak (see Fig 1 of Ref 1). The data was provided by Dr. Olga Gornik, Faculty of Pharmacy and Biochemistry, University of Zagreb, and cleaned by the GlyGen team to make it ready for machine learning algorithms. This dataset can be accessed on GlyGen's data page(https://data.glygen.org/GLY_001045).

Dataset Meta-Data

This dataset consists of 74 instances with 51 columns.

Description of the headers are as follows:

Sample: The unique identifier of the patient

Cohort: The study cohort

BL_AGE: Age at baseline

DIAB_AGE: Age at follow-up

GP*_*: