Fueling Drug Discovery with AI-ready Data: Running Automated Assay Cascades in the Digital Lab
Presented at SLAS2021 Digital
Drug discovery is an expensive, long and uncertain endeavor. Following its successive industrialization during the past three decades, it is now undergoing a digital transformation, with the goals of better leveraging scientific creativity, experience distributed across teams, and institutional knowledge while favorably tipping the cost-success balance of discovery. An important part of this process is data workflow automation, first introducing and then automating digital lab workflows.
Compound screening is the process fueling the discovery cascade in many instances, and as such has pioneered robotic process automation in small molecule discovery and increasingly also in biologics discovery. In this tutorial, we show how the corresponding data workflow is being realized and automated, enabling fast and rich assay cascades to fuel project team decisions and in-silico predictions.
We will guide you through the automation of assay registration and experiment design, data capture, data processing, QC, and analysis. We will touch on the required semantic annotation, quality assurance, standardization and experiment-adaptive business logic to produce FAIR data in the process.
The result are concise, deep bioactivity result sets, where decision-ready summaries efficiently enable the scientist to plan the next experiment, and AI-ready, fully structured and annotated multivariate data sets inform automated in-silico predictions. A corresponding case study will close the presentation.