* Exploratory Analysis of Epilepsy Biomarkers Using Untargeted Metabolomics Across Multiple Cohorts *
Authors: K. T. Oja, M. Ilisson, K. Reinson, K. Muru, T. Reimand, A.D. Kennedy, L. Sommerville, G. Michelotti, S. Pajusalu, K. Õunap
University of Tartu, Institute of Clinical Medicine, Department of Genetics and Personalized Medicine

Corresponding author: K.T. Oja

Contact information: kaisa.teele.oja@ut.ee

Address:
University of Tartu - Faculty of Medicine
Genetics and Personalized Medicine Clinic 
2 L.Puusepa street 
50406 Tartu
Estonia

*** Introductory information ***
This dataset contains the metabolomics and lipidomics data of 110 pediatric samples. It was collected for use in a paper together with data from two Estonian Biobank cohorts (access is described here: https://genomics.ut.ee/en/content/estonian-biobank). 
The pediatric cohort data is being made public both to act as supplementary data for the publication and in order for other researchers to use this data in their own work.


*** Purpose of the research ***
The goal of this study was to determine whether there are similarities in the metabolic profiles of patients with epilepsy despite different etiology, seizure frequency, seizure type, and patient age.
To accomplish this, we compared available untargeted metabolomics and lipidomics data between individuals with and without epilepsy in three cohorts (PED-C, AD-C, ELD-C). The PED-C metabolomics and lipidomics data is described below.
Files: "Data_Metabolon_PED-C.xlsx"


*** Data description ***
***** "Data_Metabolon_PED-C" *****
110 pediatric plasma samples were collected in the Tartu University Hospital between 2016-2018 and 2019-2023. Fasting was not required and the status is not known for all.
Untargeted metabolomics analysis was performed at Metabolon Inc using the Meta IMD test (four different methods of Ultra-High Performance Liquid Chromatography (UHPLC) instruments paired with Mass Spectrometry (UHPLC/ MS)). 
Detected molecules were identified by comparison to a biochemical reference library, with unique biochemical entries characterized by accurate molecular mass, including information on any adductation, in source fragmentation, and/or polymerization (typically dimers and trimers) and retention time/index on the chromatography columns. 
The results were reported as Z-scores, the number of standard deviations from the mean for each metabolite, which were calculated using a reference cohort of 866 healthy pediatric samples.

*** Data specific information ***
***** "Data_Metabolon_PED-C" *****
******* Column headings: *******
	COMPOUND - compound name
	COMP_ID - compound ID
	LIB_ID - library ID
	PED_000 - sample ID
******* Missing data *******
	Missing data is indicated with "NA"