Drug discovery has evolved over a long time, and the evolution continues - for the past several years, there has been an increased focus on data depth, data quality, and reproducibility. This is partly attributable to several key publications with concerning findings regarding reproducibility in biomedical research that have come out over the past years, but can equally be attributed to the recent rise in Artificial Intelligence and Machine Learning applications for drug discovery, with their need for large amounts of high quality data for model training and validation.
In light of these trends, it becomes clear that there is a need to address data generation at its core - by improving data quality, data depth, and data capture every step of the way, from target to hit, lead, and candidate. Arctoris has developed a robotics-enabled set of core capabilities and processes from automated cell culture to inhibitor profiling and candidate characterisation, providing an unparalleled depth of data capture, going beyond the current state-of-the-art of conventional assay setups. Exploiting gold standard assay methodologies coupled to state-of-the-art dispense capabilities and rapid analytical integration, we have demonstrated that we can deliver robust and reproducible assays within days and profiling activities within hours. Reagent characterisation, assay development, calibration, and optimisation are expedited through systematic multifactorial experimental design, facilitated using high density plate formats. Our platform affords nine orders of magnitude range in liquid volume handling, with picolitre precision and contact-free digital dispensing for true, non-serial, independent experimentation enabling experiments to be designed exclusively for each molecule. Fully automated protocols can be optimised, validated, versioned, and explicitly encoded within hours.
Using biochemical assay systems as a case study, data highlighting accelerated assay onboarding, compound profiling and mechanistic characterisation, will be shown. Extensive analysis of these will be used to highlight the importance of robustness and its relationship to inhibition thresholds used to identify hit molecules as well as informing strategies to minimise false negatives. Its value for downstream use in training machine learning algorithms or to supporting more traditional hit to lead campaigns will also be discussed. Finally, the talk will present a look along the drug discovery process, and how wet lab automation can enable the successful design and execution of more complex assays, from iPSCs to 3D cell model systems, that ensure the contextual depth of information necessary to support better decision making is available earlier in the drug discovery process, thereby delivering significant cost and time savings.