Preserving context across teams, technologies and organisations has become essential for collaboration, reproducibility and long-term research success.
As life science research organisations attempt to accelerate drug discovery and development, data continuity issues persist. Gaps in data pipelines suspend progress and waste limited resources, and hidden costs, brought by a lack of contextual scientific information, compound financial liabilities.
Researchers frequently inherit datasets without the rationale for earlier decisions. Meanwhile, inconsistent metadata, disconnected workflows, and manual data transfer cause difficulties impede progress.
Here, we take a look at the key causes of scientific continuity breakdown that occur across R&D workflows, and how a new paradigm can be established with the aid of end-to-end lab orchestration.
Siloed LIMS and ELNs
Lapses in data continuity across research projects are not a new problem. Much of this issue has stemmed from siloed laboratory information management systems (LIMS) and electronic lab notebooks (ELNs) that cannot solve continuity gaps alone.
ELNs were originally designed for compliance and recordkeeping rather than functioning as data organisation frameworks. While ELNs appear to unify workflow and decision-making support, when researchers collate data, many find ELN capabilities to be limited to compliance.
“Life science decision makers have tended to push ELNs into areas the technology was never intended for,” said Dotmatics’ Chief Strategy Officer, Michael Swartz. “It was like using SharePoint to manage your bank transactions.”
Meanwhile, LIMS software was initially utilised in manufacturing and healthcare to track sample provenance but proved too rigid for R&D workflows. Enlisted to track the provenance of samples and correlation with test results, the automation involved often becomes repetitive and leaves out the building blocks needed to secure clinical trials.
“The industry has a common habit of repeating mistakes, versus recognising when something hasn't worked and doing something different,” said Swartz. “The biggest bottleneck is essentially an innovation problem.”
Maturing the software landscape
Drugs with high molecular or target novelty have been found to outperform less novel drug discoveries by triple in revenue, in line with prescription volume. New modalities including monoclonal antibodies, antibody-drug conjugates and biospecific antibodies continue to make up the majority of projected pipeline value.
However, new modality development occurs within a biologically complex, immature software landscape. Inadequate digital representation of testing materials severely hinders the design-make-test-analyse cycle outside small molecules. For every experimental technique, there is a specialised software product to handle the execution of the experiment type. Research teams need a place to bring workflows together in one system.
"The whole idea is to optimise the properties of the material that's being tested, based on your test results. But if you don't have a good digital representation of what you're testing, you're in a world of hurt,” said Swartz.
“In most cases, apart from small molecules, even with antibodies, the digital representation is strikingly insufficient. So, researchers can often wind up with a state of affairs where they probably don't have an accurate enough picture to correctly progress.”
To fully take advantage of these value opportunities, R&D teams should consider a new paradigm that utilises an all-encompassing layer of automation and fully connects the extensive project phases at play. With help from a scientific intelligence platform, instrument data can be unlocked with ingestion and up-to-date data modelling.
Ensuring consistent terminology
Research rarely happens within a single laboratory or discipline. Projects move between biology, chemistry, computational science, screening, and development, creating numerous opportunities for valuable experimental context to be lost.
Biological and chemical staff often apply differing terminologies towards complex concepts and discoveries. With medicine discovery phases alone typically lasting around three years before trials can commence, standardisation across the project lifecycle is required to retain insight accuracy.
Research organisations can make use of ontology frameworks, either by implementing custom definitions or by utilising publicly available resources. From here, ontology management tools maintain traceability by classifying materials and sorting new entities that enter the research network.
“The semantic layer on top of things can be remarkably challenging. You need infrastructure that can provide a full digital picture of wet- and dry-lab tools,” said Swartz.
“Once you have tools that can manage those processes with high fidelity, you have a lynchpin to place ontology systems on top. This gives you a data scaffold to then hook all of your data and provides a framework for a systematic application of ontologies.”
Maintaining connectivity across workloads
Having AI-based data management in the background of R&D projects allows researchers to determine where adjustments need to be made, and where pitfalls occur. Rather than persisting with insufficient feedback loops, scientific research data can be streamlined to provide a clear picture of real-time progress.
To stay ahead of the competition, research teams need effective prevention of insight losses. This can be facilitated by a no-code platform that brings flexible configuration of workflows, tasks, data models, and governance settings. Going beyond the limitations of individual ELN and LIMS tools, end users can search across every project development in the organisation, while AI takes care of sorting and classification without the need for dataset migration.
New software solutions under development at Dotmatics, a Siemens company, represent a step change that prevents the pharmaceutical industry from having the same conversations five years in the future.
This is the second article in a four-part series exploring the importance of data continuity to drug discovery. With connected scientific data becoming the foundation of modern drug discovery innovation, investing in a platform that fits and enriches your current research workflows can allow pharmaceutical firms to stay ahead of evolving patient and organisational requirements. You can find more information below:
https://www.siemens.com/en-gb/solutions/material-research-design/
