Understanding and Classifying Metabolite Space and Metabolite-Likeness. 2011, PLoS ONE
Julio E. Peironcely, Theo Reijmers, Leon Coulier, Andreas Bender and Thomas Hankemeier.
Short Descritption Of The Paper
We have studied the chemical space of human metabolites.
We have built a computational model that assigns a metabolite-likeness score to input molecules.
We have studied which physicochemical properties and moieties are characteristic of human metabolites.
Long Description Of The Paper
This work describes the application of computational tools to predict the metabolite-likeness score of a molecule, i.e. how much it resembles a metabolite, and to acquire a global understanding of how the space of human metabolites is organized. In order to achieve this we made use of the state of the art in machine learning and the most comprehensive collection of human metabolites. To our understanding, we are the first to provide a metabolite-likeness score and to elaborate on its use for metabolomics, specifically in metabolite identification.
The use of computational models to classify metabolites and to compare them to other families of compounds has previously been published, but no prospectively validated papers have appeared. The major contribution of the current work is threefold:
- We employ and validate a variety of computational methods to not only predict if a molecule is a metabolite or not, but also to assign a metabolite-likeness score to molecules.
- Three prospective validation sets are used to demonstrate the generalizability of our models and their applicability for drug research and metabolomics.
- We investigate which physicochemical properties and chemical moieties are characteristic of the metabolite space.
We believe that our work would be of interest to medicinal chemists performing drug research, but specially to scientists in the field of metabolomics. Our aim is to show the metabolomics community how they can benefit from chemoinformatics.
Therefore, we devise further development and application of this work to accelerate discoveries concerning metabolite identification.
Calculate Metabolite-Likeness Of Your Molecules
If you would like to calculate the Metabolite-Likeness of some molecules, please contact me ( peyron #at# gmail.com)