Accelerating Assay Optimisation with Automated Design of Experiments (DoE)
Biopharmaceutical companies must optimise their assays to ensure they are robust, sensitive, specific, and precise
Assay optimisation is a sophisticated and challenging process that needs to be completed quickly to enable timely delivery of drugs and diagnostics to patients
Design of Experiments (DoE) is a powerful way of optimising assay development but it relies on automation which is difficult to implement in bioscience labs
Our software platform Antha facilitates sophisticated DoEs for assay optimisation with its codeless, user-friendly interface and robust, flexible workflows
It is estimated that an average biopharmaceutical company executes about 50 million assays per year . To ensure the safety of patients, it is essential that each of these assays is developed to be robust, sensitive, specific, and precise.
Assay optimisation can be a challenging bottleneck in drug development because condensed timelines and increasingly complex modalities make it difficult for scientists to thoroughly explore assay design space.
At Synthace, we often observe two distinct parts of the overall value-chain where assay optimisation is routinely conducted using Design of Experiments (DoE):
Within research assay development, assays are optimised using DoE prior to being run within high-throughput screens, both primary and secondary. Predictive assays and cellular models are a key area of focus for the wider industry
Later in development, analytical development teams are utilising DoE to optimise release assays, including potency and purity. DoE may often be used upstream of data-rich analytical techniques, such as HPLC or MS/MS, in order to optimise these methods for the particular biotherapeutic
DoE is a very powerful optimisation method that is becoming common in assay optimisation to efficiently explore the design space. However, DoE relies on lab automation, which is challenging to implement in bioscience labs due to lack of flexibility and the need for advanced programming skills.
In a recent live demo, Scientific Solutions Consultant Jessica Rupp revealed how our software platform Antha overcomes automation challenges and empowers scientists to easily perform sophisticated DoEs for optimising their assay development.
The Pressing Need for More Efficient Ways of Performing Design of Experiments
The DoE approach is a systematic investigation capable of elucidating not only the main factors but also the interactions between them in the same experiment. DoEs can often contain 10s to 100s of runs as well as multiple iterations of experiments in order to screen factors, refine the design space, then optimise the protocol. This process requires laboratory tools that can provide both high throughput and flexibility.
Although automated liquid handling is able to increase throughput, the inflexibility of vendor software often limits the sophistication of experiments that can be executed. For each DoE iteration, the liquid transfers involved must be calculated, planned, and programmed into a vendor-specific method, which can make the setting up of each experiment painstakingly tedious and time-consuming.
Therefore, there is a pressing need for more efficient ways of performing DoE for assay optimisation, especially when it comes to drug development and diagnostic assays. Quicker optimisation and validation of these assays would enable their timely delivery to patients.
Antha Facilitates Design of Experiments for Optimising Assay Development
We have developed a software platform, Antha, to empower biologists to use automation in their labs and therefore revolutionise their R&D. Antha overcomes the challenges of lab automation by dramatically reducing the time-cost of planning and eliminating the need for programming involved in most automated assays.
With its user-friendly interface and flexible protocols, Antha allows scientists to easily perform sophisticated DoEs to optimise their assay development.
Key capabilities and benefits of Antha include:
Automated planning & programming of sophisticated DoEs
Antha automatically performs calculations, plans and optimises liquid transfers, and generates device-specific instructions to execute DoEs on selected liquid handlers, supporting higher assay sophistication than most vendor software.
Protocol flexibility & adaptability
Workflows built in Antha can be saved and re-used for multiple iterations of DoEs by simply uploading a new design file and adjusting the experimental set-up. Unlike most automation protocols, they do not need to be re-written from scratch for each iteration.
Lowered barrier to automation use
Antha’s codeless interface, which enables seamless design and iteration of sophisticated automation workflows, is easy to use and opens up automation to scientists with little or no experience.
Antha provides full sample provenance, including the well location of each DoE run and every single liquid transfer, which ensures data reproducibility.
Antha creates robust, device-agnostic workflows that can be executed on various liquid handling devices, while retaining their integrity.
Collaboration & transparency
Antha’s secure cloud-based interface enables method and data sharing across teams and sites.
To see Antha at work, learn more about its DoE capabilities for assay optimisation, and find out how it was leveraged to design and optimise a sophisticated 7-factor 192-run DoE, watch the full demo.
How Can Antha Help with Your Assay Development?
Our demos are accompanied by live Q&A sessions, where our experts Jessica Rupp, Luke Cach, and Milena Stanković answer questions from the audience. This demo once again received a lot of great questions, a few of which are presented below:
Q: How does Antha interact with statistical software?
A: Antha users can continue to use their statistical software (JMP or DX) or a basic excel sheet to choose their factors and create their design file. This file is uploaded into Antha, which reads its, performs calculations, plans out liquid handling steps based on the conditions specified for each run, and generates device-specific instructions. The experiment is then ready for execution.
Q: Can Antha’s DoE capabilities enable machine learning applications?
A: One of the challenges with applying Machine and Deep Learning techniques is that highly contextualised and structured data streams are needed. Automated DoE is an important complement to active learning-based tools as it provides the means to iteratively explore biological design space.
In one of our case studies, we partner Antha with Lab Genius’ platform EVA, which is an autonomous AI-driven evolution engine for protein engineering. Antha’s DoE workflows were utilised by Lab Genius to plan, optimise, and execute the generation of robust DNA libraries that could be used to map the fitness landscape of new therapeutic proteins in the EVA platform.
Q: Does Antha integrate with analytical devices involved in assay development?
A: Our team is currently working on integrating Antha with a variety of devices that can be used to map samples to their resulting analytical data. These types of data workflows can vary based on the use case. If you and your team are working on this, please contact us at firstname.lastname@example.org to discuss how we might be able to help you.
This demo exhibited one of the many ways Antha can be used in analytical development to make automation more open, flexible, and extensible. To talk to us directly about your automation needs (DoE or otherwise), please email us at email@example.com.
Machine Learning: It’s All About the Data
How to build a strong data foundation for machine learning applications.
Synthace Unveils First Life Sciences R&D Cloud Addressing Complexity, Speed & Reproducibility for Scientists
First ever no-code platform lowers barriers to automated biological experimentation and insight sharing.