Welcome to MZedDB the Aberystwyth University High Resolution Mass Spectrometry Laboratory database for metabolite signal annotation.

The content of MZedDB is derived largely from established repositories of curated information relating to metabolites. The databases accessed are primarily those integrated directly with either genomics and/or metabolic pathway data. Information in ‘super’ repositories such as PubChem and ChemSpider, which contain many entries relating to chemistry of non-biological origin, is accessed by integrated secondary search routines. Molecular information is reprocessed so that each identical structure constitutes a single entry in the MZedDB database:
  • an identical structure has exactly the same description of skeleton and stereochemistry. The same metabolite may have several entries if stereochemistry is partially (or not) described. The name of the entry is taken arbitrarily from one of the given synonyms in each original database.
  • chemical structure processing consists of removing salts (i.e. keeping the largest component), molecules with less than 6 atoms or exotic elements. When possible in an automatic fashion, charged entities have been converted to neutral compounds by addition or removal of hydrogen. In this respect, there might be slight differences in molecular structure representation in the original database and the one used in MZedDB. In some instances automated conversion could not be achieved.

Note that no manual curation is performed so that if information relating to a molecular structure is flawed in its original database, the error will be retained in MZedDB, but hopefully the same compound will have been correctly described in the other database. Automatic updates to extend Metabolite Search coverage take place approximately every three months.

Our main interest lies in the automated, high throughput analysis of data derived from ‘soft’ ionisation mass spectrometry methods (e.g. electrospray) applied to crude extracts of biological samples. MZedDB works on the principle that many ionisation products will represent adducts, isotopes and neutral loss fragments of parent metabolites based on a specific set of generalised physical rules. With this in mind a Putative Ionisation Product PIP Search function is provided for the putative annotation of m/z signals which can be parameterized depending on the biological matrix under study, extraction solvent and HPLC mobile phase. With the advent of more accurate MS instruments, tools to identify molecules based on chemical formula predictions Molecular Formula Generator have been added. As much as possible, information is transparent, results can be saved or exported and the system can be easily hacked to automate fastidious repetitive queries. Based on either a predicted molecular formula to an actual MZedDB entry it is possible to generate possible adducts (M->IP) using the Adduct Manipulation tool, with or without applying rules. The reverse transformation (IP->M) can be performed so that given the molecular formula of a potential ionisation product, all possible formula of the parent compound can be enumerated. Automated searching of MZedDB can be achieved directly in an R environment using the function R<->MZedDB. Input data can be whole matrices or a list of explanatory m/z after data mining.

Acknowledgements: Thanks to those who have made their efforts available to others.

This is a freely-accessible service that uses information available elsewhere. However, we request our visitors to use the service with parsimony. Heavy trafficking or downloading could result in access from a particular IP address or domain being denied. We cannot guarantee that the service is error/bug free and we welcome your feedback. We are not held responsible for any misuse of the information provided.

The information provided on these web pages, is the responsibility of the High Resolution Mass Spectrometry Laboratory headed by Prof. John Draper and not that of the Aberystwyth University (AU). Any opinions expressed here are those of Prof. John Draper & Co-workers and are in no way to be taken as those of AU.