Insights

Unleashing the Power of Digitalisation

What’s so important about digitalisation anyway?

Katrina Costa, MA (Oxon), MSc
Science Writer, Open Pharma Research Ltd. December 2020

Yes, life science companies are embracing the power of digital. There is growing realisation that archaic R&D processes need to be disrupted and reengineered in order to deliver much needed medicines for unmet patient needs. However, digitalisation is about more than just creating data. It’s about not wasting time and resources on trying to access the data, avoiding repeating experiments and ensuring the data is readily available and re-usable.

Digital transformation is even more radical – it involves a greater focus on patients and overhauling the use of digital technologies to create a more unified and connected data lake.

The goal of the Lab of the Future is to drive greater scientific innovation, and ultimately get the right medicines to patients, faster.

However, unlike our Silicon Valley counterparts, we have well known legacy challenges relating to the adoption of new technologies and working practices. Are we as an industry truly ready for digital?

Unlocking the power of complex life science data

The potential benefits of digital transformation include:

  • Faster drug discovery and development
  • Greater insights through machine actionable data
  • Streamlined regulatory approvals
  • Operational excellence in manufacturing

For the life science industry to succeed in its digital transformation journey, data needs to meet FAIR standards all along the value chain, laboratory instruments need to be well-integrated, and sufficient resources and expertise dedicated to realise this goal.

What are the challenges to digitalisation?

At an executive level, there is significant support for digital transformation. However, data practitioners and users are facing issues of low data quality, lack of standards and poor data linkage. These present numerous challenges to pharmaceutical and technology companies when moving towards effective digital transformation. Other obstacles include lack of expertise, low level of buy-in from scientists and issues with data not being machine actionable. However, a recent Lab of the Future survey revealed that Data Governance was the biggest single issue in pharma digitalisation (see Digital Dialogues in ‘further reading’).

Poor data quality and data governance are not just limited to IT or Data Transfer, it impacts departments across the organisation including:

  • R&D
  • Quality Assurance
  • Regulatory Affairs

Another big problem with poor data governance is the risk of losing supporting data that have been discarded or lost in silos.

The Roche and Merck Group Experience

For successful digital transformation, it’s vital that data are high quality, as well as Findable, Accessible, Interoperable and Reusable (the FAIR standards). Additionally, the data needs to be machine-actionable, requiring minimal human intervention. PricewaterhouseCoopers have estimated that the minimum annual cost of not having FAIR data is €10.2 billion. 44% of this cost waste is in the creation and collation of data, with absent or poor metadata. This leads to high levels of data cleansing, often by data scientists, further down the pipeline.

To address these issues, Roche have adopted FAIRification – generating high quality data at every stage of the value chain. This has greatly improved the efficiency of their R&D process because, as the data is FAIR from the point of entry, it doesn’t require cleansing at later stages.

Roche use a ‘Data Commons’ architecture framework to assist with FAIRification and data management, with a focus on harmonised data access points and integrated data sets. All data is annotated with rich metadata using a terminology management system.

Roche also treats data quality as a service, and uses a list of data quality KPIs, such as completeness, correctness and consistency. It’s also important to remember that just because data are high quality doesn’t mean they are FAIR, and in this instance we could lose access to good data. Conversely, FAIR data isn’t necessarily high quality.

To address the difficulties of data governance, companies need a cultural change and corporate-wide strategy to achieve successful digital transformation. To this end, Roche have created a ‘FAIR data playbook’ for Data Managers, and are also developing one for Solution Architects, Business Analysts and Software Developers, to ensure FAIR is at the heart of everything they do.

Connecting instruments in the lab

Roche recognise the importance of laboratory instruments being connected – both within individual laboratories and across the organisation’s many silos.

Benefits of instrument integration include:

  • Improving operational efficiency
    (time-saving, speed up processes, reduce repetition of experiments)
  • Improving quality, integrity and compliance
    (remove error prone manual processes)
  • Capitalising on knowledge
    (data sharing, mining and analysis, wide collaborations)
  • Simplifying the IT landscape
    (save time and use a single system)

Roche’s goal was to get the data out of silos and into a data lake in a standard format, ensuring the R&D process would be far more efficient. They couldn’t find a vendor to suit their needs in the R&D space, so they developed the Roche Product Technical Development (PTD) instrument integration platform. This platform is essentially a middle layer – acquisition software collects the instrument data, which is transferred to their platform, and then uploaded into the cloud.

Benefits of their platform include:

  • One platform for everything
  • Aiming for 100% FAIR
  • Standard data components combined
  • No vendor lock-in
  • Data available in near real-time
  • No licence costs

 

A tool for enhancing lab data

Merck Group recognises that a big challenge to FAIR data is losing data and missing opportunities for secondary usage. Merck Healthcare adopted BSSN converters – software that collects data from laboratory instruments and converts it into an open AnIML (human and machine-readable) format. They created a storage system, visualiser, data processor and analyser. After achieving this, BSSN won the 2019 Frost & Sullivan award for Technology and Innovation, and Merck Life Science have now acquired the software and committed to the open standards for the life science industry. They want to provide a technology that is usable for a broader audience. This vendor-neutral lab solution will help harmonise the data lake and make it more structured, integrated and re-usable.

Where next?

In the Lab of the Future, digital transformation will bring huge benefits to patients. In fact Roche have already experienced some early wins, as Dr. Martin Romacker, Senior Principal Scientist at the organisation explained: “We’ve reduced the time it takes to get drugs to the market and medicines to patients”. Savings in time and processes also means less waste.

GSK have also seen improvements in the drug discovery phase, which is less subject to regulation. GSK’s Tech Director and Senior Product Owner, Dr. Penny Smee, said: “Our data is being captured in a more structured manner, which means it’s then more readily available for us to re-use it, to apply the AI/ML elements, and get more from the data we have”.

Dr. Smee also believes digital transformation holds great promise for the manufacturing phase and ensuring the data comply with safety regulations. “We spend months and months of staff time collating data, transcription checking, making sure it’s right – manually. Imagine the power of being able to push a button and say ‘create me the table for the regulatory file’.”

However, data transformation is still at an early stage. Dr. Haydn Boehm, Head of Commercial Marketing – Connected Lab, Merck Group, said: “One issue is the quality of the data coming out of various systems. This is your base of the pyramid that allows loftier goals – for example, AI is at the top of the pyramid. Your modeling programmes are based on the data you have. But things are getting better all the time. If you can first govern and optimise your data at the source, we’ll see progress a lot faster. It’s quality in, quality out.”

Other ‘top of the pyramid’ goals include machine learning, non-siloed collaborative access, the move to precision medicine… the list goes on.

Clearly, digitalisation is not just important, it is critical in delivering the Lab of the Future and ultimately in creating the precision medicines of tomorrow. The path towards digitalisation will not be easy. Obvious hurdles need to be overcome in order to align data, technology, people and processes. However, as the current global health crisis has revealed, the urgency for getting medicines to patients faster has never been more pressing – it’s time to tackle the obstacles to digital transformation.

Further reading

  1. ‘The FAIR Toolkit for Life Science Industry’, Pistoia Alliance. https://fairtoolkit.pistoiaalliance.org/
  2. ‘How structured and accessible research data can drive digital transformation in the lab’. (26 October 2020). Digital Dialogues session, Lab of the Future. https://www.lab-of-the-future.com/digital-dialogues/