As a company driven by AI and designed to be intrinsically digital, Generate Biomedicines has worked towards its lab-of-the-future vision since its inception. Our platform combines aggressive automation to realize the promise of de novo protein design, synthesis, and optimization, with the necessary configurability to accommodate emerging scientific and technological innovations. Our software and hardware systems are designed to close the design-make-test-analyze loop with as much flexibility and as little friction as possible. In this loop, lab automation enables data to become matter during synthesis, and then converted back to data during measurement. Our informatics toolset is anchored on a FAIR database that enables all measurements and metadata collected throughout the entire process to be easily captured, recorded, and retrieved to drive informed and timely decision-making at all levels. Fit-for-purpose metrics, KPIs, and trends delivered through interactive data visualizations enable monitoring the effectiveness and efficiency of our end-to-end discovery process, and the iterative training of AI/ML models to drive both protein design and operational improvements, thus allowing us to optimize drug candidates better, faster, and cheaper than any competing platform. Our embodiment of the lab-of-the-future is supported by a cross-disciplinary team of robotics, software, and data engineers and data scientists working hand-in-hand with R&D scientists in a highly collaborative and dynamic environment fueled by an empowering culture and a shared, audaciously ambitious vision.

With the emergence of Platforms-as-a-Service such as Lifebit's CloudOS, DNANexus St. Jude Cloud, Broad's Terra, Velsera's Cavatica, and others, we have the opportunity to focus on more science. Modern Workflow languages (Nextflow, CWL, and WDL) and containerization at the process level, along with the use of these PaaS's, makes cloud-based data and cloud compute accessible for everyone.

The presentation will trace the evolution of the Brigham and Women's Hospital Clinical Laboratories over the past two decades, including the provision of HIS-embedded, physician order-entry based clinical decision support tools, core laboratory automation, and growth of advanced diagnostics capabilities and programs.

The Instrument Digital Twin has been designed as a virtual representation of all R&ED instruments at BMS, spanning their lifecycle, updated with real-time data, including instrument operational state and all raw data generated on the physical instrument. The Digital Twin is physically implemented as a cloud-based data store; it has an instrument ontology, a composable architecture, data ingestion capabilities, and exposes its content via a set of APIs. The Digital Twin is generic, extensible, vendor-neutral, and enables integration into various downstream applications. BioBright DarwinSync is the core laboratory data acquisition and analysis platform that feeds into the Digital Twin. The Instrument Portal is a user-centric front-end to the Digital Twin and enables viewing and consuming the content of the Digital Twin via the APIs and is the one-stop shop for Digital Twin capabilities. The Digital Twin replaces the legacy approach to laboratory data acquisition and analysis.

Regeneron's Research and Preclinical Development organization invests millions of dollars each year to acquire and operate specialized lab instruments that support scientists in the discovery of potential new medicines. To optimize the usage of these instruments across finite lab space, a project was undertaken to measure instrument utilization. We sought to answer questions like: Are there opportunities to better utilize existing instruments? and Could we plan and budget more effectively if we better understood existing instrument usage? Additionally, this data enables new opportunities to optimize instrument maintenance and other lab processes to increase the likelihood Regeneron is achieving the most value from its investments. Regeneron is now measuring utilization across 600 of the most expensive and/or lab space constrained instruments and using this data to make informed decisions to change processes or defer capital purchases that optimize utilization of a given instrument.

HigherSteaks is a lab-grown meat company on a mission to reduce traditional meat consumption by having a 5% share of the global pork market by 2030. HigherSteaks is ushering in a new era of sustainable food production through lab-grown meat and is doing so through design of experiments (DOE) to accelerate media optimization, and by running factor screens based on different cell lines of porcine stem cells. Synthace, a life sciences SaaS platform, and HigherSteaks’ core experimental solution, allows them to use DOE and fractional factorials—screening 22 factors, as well as interaction profiles, in only 320 experimental runs. Synthace saved HigherSteaks 6-9 months of work and delivered an optimized media for pig stem cells 15x cheaper compared to commercial media. This experiment proves the concept, and demonstrates the viability, of DOE methodologies for the cultured meat sector, without additional lab investment.

Digitalization is recognized as a key enabler in pharmaceutical R&D, yet the fully digital lab is in most places still a vision on paper and slides. The talk will summarize key digital capabilities most labs still need to evolve further, discuss realistic timelines to reach fully digital ways of working within and across labs, and present opportunities to speed up the path towards the vision of a fully digital lab on the example of R&D labs in Roche Pharma Research and Development.

Data escape from the wet lab via different modalities, ranging from sneakernet to automated data capture systems. We will present a decision tree to help wet lab scientists and IT decide how to get raw data from the laboratory into computational spaces in ways that support FAIR principles.

Successful FAIRification of lab data presents well-known challenges due to the nature of experimental workflows. We will present recent efforts in AstraZeneca early research to Digitize in lab workflows. We will discuss our work from policy, software selection, and solution implementation including obstacles we have faced.