Category: Data Analysis and Informatics

1263-D - An Informatics Workflow for Amgen’s Affinity Mass Spectrometry Platform.

Tuesday, February 6, 2018
5:00 PM - 6:00 PM

  Amgen’s affinity mass spectrometry screens assess the binding of small molecules to a target protein. By determining small molecule relative abundance, with and without protein present, the percentage of sample enrichment (PoE) can be calculated. The Agilent Rapidfire® mass spectrometer (Quadrupole Time of Flight) enables compound pooling such that a library of 500K lead compounds can be condensed to over 200 compounds per well in less than eight 384 well plates, dramatically reducing the amount of protein consumed.

  Here we describe an informatics workflow for such at-scale Affinity-MS based screening. Batch analysis is made possible by custom software that automates the assembly of an Agilent MassHunter® worklist to be run with Agilent DA Reprocessor®. A single MassHunter Method is applied to each of the pooled samples with well-specific mass and molecular formula databases referenced to avoid false positives. Considering that there are ~5000 sample wells across with and without protein modes, this would be an almost impossible manual process to achieve robustly and in reasonable time.

  Results are then imported into Genedata Screener® where a custom script-parser implementation de-convolutes the pooled wells into virtual plates. The Screener® Experiment model utilizes free formula calculations to determine the Percent of Enrichment (PoE) based on sample error flags and the background rate of detection. Additional QC and validation is performed via a custom Screener® High Content Screening Plugin implementation that displays compound structures and the chromatograms and mass spectra of the with and without protein samples. Results are then reported to Amgen’s Assay Data Warehouse.

Mark Southern

Senior Engineer
Newbury Park, CA

Experienced professional with 20 years in the scientific informatics field covering the pharmaceutical, biotech and academic sectors. I have been at Amgen since 2015 working as a software engineer within Discovery Technologies, specifically on the UHTS platform. Previously I was with The Scripps Research Institute during the entire Molecular Libraries Initiative from the NIH Roadmap, the MLPCN grant being the largest ever awarded to the institution. I helped grow a lot of the production scale operational informatics infrastructure for the Scripps Research Institute Molecular Screening Center during the pilot phase (2005-2008) of the project when I was a Senior Systems Analyst and then ran the Informatics group as Associate Director, overseeing screening informatics and cheminformatics during the production phase ( 2008-2014).