Digital: Digital Transformations

Pharma laboratories should be looking to innovative digital techniques such as data, AI and the cloud to aid drug discovery and research

It can take 10-15 years and cost in the region of $2.5bn-3bn to take a drug from initial concept to commercial availability.1 Added to that, only about one in ten drug candidates make it to regulatory approvals. With costs and odds like these, organisations across the life sciences industry must look to technology to accelerate timelines, reduce costs and increase the likelihood of successful drug candidates making it to regulatory approvals.

Whether you are a clinical diagnostics facility looking for faster turnaround times, or a pharmaceutical research lab looking to speed up the drug discovery pipeline, you will no doubt be using technology across multiple areas of the operation.

But, are you going fast enough or deep enough with your digital ambitions, or are we merely scratching the surface of what data, cloud and artificial intelligence (AI) can achieve in revolutionising drug discovery and development?

Digital transformation is the process of using technology to continually evolve all aspects of a business. For pharma and other research organisations, this involves implementing new technologies to reshape lab workflows, expand collaboration, maximise operational efficiency, modernise experimental methods and transform the way scientists record and analyse data. It’s not a ‘one and done’ tactic; it’s a strategic process that demands a continuous focus on the areas that need improvement and what technologies can have the most impact.

Every lab will have unique challenges and goals, often best understood by the researchers themselves. The scale of digital transformation is also unique to each research or testing environment. Are you looking at a whole-scale renewal of the processes that underpin the business and research efforts, or a more targeted project to solve a specific issue or research challenge?

Whatever the aim, scale, or challenge being addressed, the vast majority of digital transformation projects will need to address three core areas: data; cloud; and AI.

The amount of data produced in a lab will be directly influenced by the nature of the organisation’s research. For example, a company focused on personalised cell therapies can produce vast amounts of data due to its unique R&D capabilities. These types of companies anticipate a continual rise in the amount of data produced as the demand for cell therapies rises whilst continuing to meet compliance guidelines. This requires recording the handling, processing and testing of cell therapy products, in turn generating huge amounts of data.2

As a company like this grows to a mid-size clinical biotech, data types will also grow, expanding to include whole genome sequencing, RNA and single-cell RNA sequencing, single nucleotide polymorphism arrays, multi-omics, histology and other wet lab data sets, microscopy, and imported public data sets.

In addition to the volume of data, it is not uncommon for there to be many different data sources within a lab setting. From the genetic data produced by next-gen sequencers, to images obtained through microscopy and X-rays, this vastly varied information is often siloed in different repositories or different formats. This creates a range of challenges for the scientist, including inefficiencies in accessing existing data, limited cross-referencing abilities across data sources and fragmented insights due to the inability to combine data sources effectively.

These examples demonstrate how the creation of data will continue to accelerate, and the types of data required across increasingly multi-omics research will continue to grow. Because of this, how a business manages, secures and integrates data should be an ongoing focus for any digital transformation strategy.

Implementing a lab information management system (LIMS) makes data more readily accessible, allowing researchers to retrieve, analyse and compare data quickly. Integrating data from various sources, including device and instrument data, for example, gives researchers a more holistic and comprehensive view, ultimately allowing them to get the most out of their analysis and overall investigation.

Labs transitioning to centralised data management systems, as a discrete project or part of a wider digital transformation strategy, should be aware of the challenges. These can include data migration and compatibility issues when transferring existing data from legacy systems, as well as potential disruptions to workflows during the implementation process.

The upside of deploying a solution that enables the integration of all lab data cannot be underestimated. By ensuring data collection is accurate, reliable and accessible, integrated data management enables researchers to quickly retrieve, analyse and compare data. By integrating data from various sources, researchers can also gain a more comprehensive understanding of the data, resulting in deeper insights.

The life sciences industry, pharma included, has historically relied on ‘on-premises’ computing infrastructure. Closed, often proprietary, computer systems provided the companies with a seemingly super-secure way to store, analyse and interrogate scientific data. The cloud was viewed with a certain amount of scepticism – putting billions of dollars’ worth of research data into the hands of a third party or sharing computing power with a competitor was seen as a ‘no-go’ area. Today, the cloud is the only practical solution. As the volume of information and data related to drug discovery has exploded, the industry is seeing the cloud as the only way to meet data storage and security, innovation and collaboration requirements.

Historically, and incorrectly, seen as less secure than on-premises solutions, the cloud offers cutting-edge encryption technologies for secure data transmission and access. These systems also utilise access controls, standby database synchronisation, backups and other features to further protect lab data, providing a more comprehensive and reliable defence against potential security threats than an on-premises approach.

Of course, sensitive or proprietary information needs to be stored in a way that ensures compliance with regulatory frameworks like the General Data Protection Regulation (GDPR), the Health Insurance Portability and Accountability Act (HIPAA) or the Health Information Technology for Economic and Clinical Health (HITECH) Act.3,4,5

A further benefit of a cloud-based approach is that regulatory compliance should be managed by the cloud provider, removing one time-consuming element from the lab or research organisation’s IT and compliance teams.

However, labs shouldn’t be limited to just cloud-based data management in their digital transformation ambitions. The biggest benefits for scientists come from the adoption of software-as-a-service (SaaS) lab informatics applications. These applications, running in the cloud, bring all the benefits of cloud storage, security, scalability, flexibility and accessibility, but they also provide scientists with the most up-to-date technologies and innovations available. In the life sciences industry, this is critically important. The tools and software used by scientists need to reflect the latest scientific thinking, not the scientific thinking when the software was first written.

It is commonplace for different labs in different locations and with different specialisms to collaborate on the same project and share critical insights. However, for this to be effective, researchers and labs must continuously share their most recent data sets with others involved in the project to ensure that the project is based on reliable, up-to-date information. A cloud-first approach to digital transformation enables research organisations to make data available across the entire organisation, as well as opening up the data to research partners and collaborators, albeit in a secure, managed and permission-based way.

Adoption of the cloud as part of a digital transformation strategy offers so much more than just ‘off-site data storage’. It provides pharmaceutical researchers with ways to look at data and information in new and novel ways, and tear down data silos that have impeded the speed at which scientists can get results.

Companies across the life sciences industry have long recognised the value of data analytics to extract valuable insights from data sets. When analysed effectively, this data can impact all areas of the business, advance scientific research, impact new treatments or influence how personalised medicine and care are delivered.

Technology has been used to assist in the analysis of scientific and experimental data for a while. From statistical analysis and data mining, to bioinformatics and systems biology, the different analytical models used in life sciences are impressive. While these data analytics techniques play an invaluable role in the drug discovery process, the industry continues to look to technology to provide systems that are easier to use and faster at providing the insights they need to drive innovation.

However, are you going far enough with your ambitions?

AI has the potential to revolutionise the use of data in labs. AI algorithms are starting to be given access to specific, often proprietary, life sciences data through initiatives like NVIDIA’s BioNeMo.6 By training the AI on specific scientific data collected from a variety of sources and data types, the results, insights and analysis are considerably more effective and reliable. These advances mean we will start to see more ‘science-aware’ AI, resulting in a real shift in how scientists and researchers interact with and use the technology at their disposal.

Like all digital transformation initiatives, AI needs to demonstrate real, tangible benefits – generally productivity gains or cost savings for the organisation. A recent 2024 Gartner Hype Cycle revealed that generative AI (genAI) has passed the ‘peak of inflated expectations’ and is now sliding down into the ‘trough of disillusionment.’7 It seems that the hype behind offerings such as OpenAI’s ChatGPT or Anthropic’s Claude has not translated into a solid return on investment (ROI) for most companies.

However, a focus on ROI is likely to lead organisations to look for ways to use AI that offer more solid potential for productivity and efficiency gains, such as AI agents that collect and use data to perform self-determined tasks.

When coupled with natural language processing (NLP), AI is set to revolutionise how researchers and lab technicians interact with the technology in the lab, effectively changing the dynamic between humans and machines. Advanced NLP capabilities – incorporating both conversational text and voice-driven interfaces – will enable users to interact with complex systems using ‘natural’ commands, making the technology more accessible and user-centric.

In the life sciences and pharma worlds, this will take the form of virtual lab assistants. Science-aware AI agents, trained on data from across the organisation and the wider science ecosystem, are underpinned by NLP capabilities. These virtual lab assistants will aid scientists in any number of tasks, from finding data and results, to creating experiments or retro-synthesising molecules. These AI-driven assistants have been trained to understand complex scientific queries, interpret contextual information and execute sophisticated tasks based on researchers’ conversational inputs.

Digital transformation is a vital and hugely beneficial journey to embark on. Across a whole host of industries, there have been radical improvements in productivity and efficiency through the adoption of digital technology. Revisiting the definition of digital transformation used earlier, it shows that it is a process, not a specific destination; a process where an organisation uses the latest technology to continually evolve. In some areas, the life sciences industries have been at the forefront of leveraging technology to improve processes and increase efficiency, but have we gone far enough? The key areas of data, cloud and AI are continually evolving, and the industry needs to stay ahead of the curve.

The research world is embarking on the next ‘stage’ of digital transformation. By improving data integration, data management and collaboration, and adopting technologies like science-aware AI, NLP and AI agents, the life sciences industry will see huge productivity gains. This will give scientists more time to do what they do best and develop innovative and creative approaches to some of the most difficult questions faced today.

References

1. Visit: phrma.org/policy-issues/Research-and-Development-Policy-Framework
2. Jannessen WE (2008), ‘Datamanagement in the cell therapy productionfacility: the batchprocessrecord (BPR)’, Cytotherapy 10(3), pp227–237
3. Visit: commission.europa.eu/law/law-topic/data-protection/data-protection-eu_en
4. Visit: aspe.hhs.gov/reports/health-insurance-portabilityaccountability-act-1996
5. Visit: hhs.gov/hipaa/for-professionals/special-topics/hitechact-enforcement-interim-final-rule/index.html
6. Visit: nvidianews.nvidia.com/news/nvidia-launches-largelanguage-model-cloud-services-to-advance-ai-and-digitalbiology
7. Visit: gartner.com/en/newsroom/press-releases/2024-0821-gartner-2024-hype-cycle-for-emerging-technologieshighlights-developer-productivity-total-experience-ai-andsecurity