Home
Home
-
National Metabolomics Data Repository
Upload and Manage
StudiesBrowse and Search
StudiesAnalyze
StudiesAs of 06/05/23 a total of 2641 studies have been processed by the National Metabolomics Data Repository (NMDR). There are 2222 publicly available studies and the remainder (419) will be made available subject to their embargo dates.
Recently released studies on NMDR
ST001474 - Metabolomics of lung injury after allogeneic hematopoietic cell transplantation - Spleen ICMS; Mus musculus; University of Kentucky
ST002697 - Employing Hindlimb Unloading Model for The Identification of Serum Biomarkers Associated with Cardiovascular and Skeletal Muscle Deconditioning.; Mus musculus; Sharjah Institute for Medical Research
ST002175 - Effect of external high-dose rate radiation on mouse biofluid metabolomic signatures; Mus musculus; Georgetown University
-
Core structures in RefMet classification
RefMet core structures (March 1, 2023)
Browse and search core structures associated with the RefMet classification system. For example, what are "Flavones", "Flavanols", "Flavonols" and "Flavanones" and what's the difference? The RefMet classification hierarchy has recently been updated to place more emphasis on biosynthetic considerations for Alkaloid, Polyketide and Prenol lipid super classes.
-
Molecular structure similarity analysis
Create molecular structure similarity networks (February 3, 2023)
This Structure similarity network tool creates a network map from a list of metabolite names (up to 500) by selecting a fingerprint type (MACCSkeys, Chem.RDK, Topological, Morgan,MorganBitVector) and similarity method (Tanimoto, Dice) with a similarity coefficient cutoff. This feature is also implemented for each NMDR study containing named metabolites (in 'Perform statisctical analysis' section). This application uses the Python-based Rdkit.
-
Searching untargeted LC-MS data
Searching untargeted LC-MS data on the Workbench (December 13, 2022)
This portal searches over 4.5 million m/z,retention time features from over 890 NMDR studies and over 1500 LC-MS analyses. Search with a m/z value and tolerance window and optionally specify a retention time value and tolerance window to restrict the search. Limit search to studies by sample source and/or species, and also by chromatography type, MS instrument and polarity. Features that have been identified by submitters will appear in the "Name" column in the results table.
-
Correlated network graphs in NMDR
Correlated network graphs using Debiased Sparse Partial Correlation (DSPC)
The Metabolomics Workbench has released a new graphical tool for estimating and visualizing partial correlation networks in NMDR studies. It uses the Debiased Sparse Partial Correlation algorithm (DSPC) developed at U.Michigan. Nodes may be mapped to chemical classification or fold-change. Study example: See "Perform Network analysis on correlated metabolites" links here
-
Lipid Notation in RefMet and lipid m/z calculation tools
View table of over 170 revised lipid abbreviations covered by RefMet, including structure examples and m/z calculation tools for a variety of adducts.
-
Convert your metabolite name to standardized nomenclature via RefMet
-
Exemplary Studies
A list of exemplary studies are listed here which adhere to the submission guidelines of Metabolomics Workbench. Specifically, publically available studies having all or most of the features below were identified as exemplary studies.
- Well-written study summary
- Detailed metadata for collection/treatment/chromatography/MS/NMR, etc.
- Post-processing details
- Presence of control samples
- Raw data availability for samples and controls
- One-to-one mapping of sample names to raw data file name
- Internal standards (with measurements)
- Clear and organized metabolite annotations
These include different analysis (GC-MS, LC-MS, NMR) and species type. We recommend looking at these studies as a model example before submitting to Metabolomics Workbench.
-
NMDR studies and Jupyter Notebooks
Analyze Workbench studies via Python-based Jupyter Notebooks. Launch notebooks on Binder or download notebooks from GitHub and run them locally.
