Overcoming data challenges in automated mass spectrometry-based screening of entire compound libraries
ASMS Reboot, Virtual Conference
May 14, 2020
Mass spectrometry is a label-free, broadly applicable technology for screening. Innovations in sampling—such as acoustic sample deposition—enable screening of entire compound libraries in a matter of hours or days. However, the analysis of the resulting data poses several challenges. Ideally, a data processing and analysis platform would provide automated identification and quantification of the relevant end points, even in complex samples; proper handling of noise in the data and assay artifacts at scale; calculation and visualization of the results of the entire screen, including the pharmacological end-points (e.g. IC50); and integration with the remaining IT environment of the organization. We demonstrate a data analysis solution that automates high-throughput MS data processing at scale and directly tackles the challenges listed above.
Raw spectra were acquired for an entire dose-response screen (hundreds of plates, thousands of compounds, 10 concentrations) using acoustic sampling coupled to mass spectrometry. The resulting data went through a fully automated data processing pipeline using the Genedata Biopharma Platform: data processing steps included noise reduction, peak detection and clustering, comparison with a list of expected target masses, and quantification. This was followed by normalization based on plate controls, exclusion of failed wells, dose-response curve fitting, and classification of the compounds by efficacy and potency. Subsequently, interactive visualizers were used to ensure data quality. Finally, the results were seamlessly transferred to the assay data warehouse.
The entire data processing pipeline ran completely unattended in a matter of minutes. Once loaded into a dedicated software platform (Genedata Screener®, Genedata) the combination of heatmaps of plates from the entire assay, coupled with visualization of dose-response curves, resulted in highly efficient quality control of the results. The configuration and parametrization of the entire pipeline using a graphical user interface enabled flexible adaptation to a wide range of assay setups. It also provided a way to leverage additional metadata not present in the raw spectra—for example, an external list of target masses—to perform analyses.
Automated analysis of data from high-throughput mass spectrometry-based assays, enabling adoption of this technology to screen entire compound libraries.