Machine Learning Supported Targeted Metabolomics

Abstract

The innovative medicines initiative funded RESOLUTE consortium is a public-private partnership systematically investigating the solute carrier family. Solute carrier transporters (SLCs) are an understudied but therapeutically relevant protein family with 446 members. SLCs regulate metabolism directly by controlling metabolite transport across membranes at the cell surface and between sub-cellular compartments. We use targeted metabolomics to study metabolism in a systematic family-wide manner and relate metabolic phenotypes to SLC function. The goal is to address SLCs individually and using a guilt-by-association principle.

We have developed a high-through-put workflow to grow inducible SLC overexpressing cell lines, extract metabolites and quantify them in a reproducible manner. We use an ion-pairing, reversed phase chromatographic separation coupled to triple quadrupole dynamic MRM mass spectrometry to allow absolute quantification of 200 metabolites of which around 140 can robustly be detected in our cellular systems representing more than 50 metabolic pathways.

After manually integrating already more than 300'000 chromatographic peaks, we are using machine learning to automatize peak picking and peak integration. Data is visualized using an in-house data pipeline. Metabolite changes in response to over-expression of specific SLCs are used to identify primary substrates directly or to deorphanize SLCs based on similarity to SLCs with known substrates. Targeted metabolomics data is eventually integrated with proteomics and transcriptomics data to further deepen the understanding of SLC biology and function.

Date
Location
Unlocking Transporters for Drug Discovery, Palais Niederösterreich, Vienna, Austria
Next
Previous