Bridging the automation gap: Challenges of data processing and analysis

Satnam Surae, chief product officer, Aigenpulse examines how automation can benefit the pharma industry when it comes to data processing and analytics.

The benefits of automation to laboratory analysis have been widely reported over the last decade. As such, the total laboratory automation market is thriving, having been valued at USD 4,315.5 million in 2019, and expected to reach a value of USD 6,301.8 billion by 2025, at a CAGR of 6.6% from 2020–2025.¹ As technology innovation progresses at a rapid pace, laboratory automation will continue to permeate a broad range of sectors, from genomics and proteomics to drug discovery and clinical diagnostics, for the foreseeable future.

Increasing throughput, reducing human error, boosting productivity, and saving time and money, for example, are what has cemented automation within pharma and biopharma R&D. Another advantage includes enhanced standardisation for accreditation and certification, where increased accuracy and repeatability throughout laboratory processes – enabled by automating operations – facilitate standardisation, thus ensuring compliance with certification and accreditation procedures.

However, automation leads to a spiralling amount of data produced by an average R&D laboratory. Data analysis is often still highly manual and requires specialist programming and data science skills. In order for R&D labs to gain maximum value from data in a realistic time-frame, data processing and analysis technology must keep pace with automation innovation.

Optimising data processing

Although significant progress has been made in dealing with the output of highly automated laboratories, true data integration platforms are still needed. The data and sample management systems currently available have provided significant benefits to the lab, but began as distinct systems and therefore created separate data repositories that require an interface or middleware to enable data to be shared.² The challenge of data harmonisation, data connection, and lack of high quality data therefore remains. In addition, there are compounding effects of errors produced by scientists manually handling data, and an inability to retrace steps from results, which is vital in drug development where each step of a process must be available for scrutiny. Solving these challenges provides huge opportunities to transform the way in which new therapeutics are identified, validated, and developed. 

A new approach is needed

There is a clear need in pharmaceutical and biopharmaceutical development for automated data processing and analysis. A platform that stores and manages all data, pipelines and outputs would not only improve data quality, but help organisations drive towards FAIR data principles. These principles, which aim to make data Findable, Accessible, Interoperable and Reusable, provide guidance for scientific data management and stewardship and are relevant to all stakeholders that directly address data producers and data publishers to promote maximum use of research data.³ 

New dedicated platforms for pharmaceutical R&D data processing are now emerging that tackle these challenges. Currently, data silos in pharma prevent the latest advancements in machine learning (ML) and data analytics from realising their full potential. By unifying silos and promoting data re-use, new ML-driven platforms enable analysts to collate information from disparate sources and easily generate meaningful insights. High quality outputs can be generated at multiple stages of the R&D life cycle and democratised throughout the organisation.

Focus on flow cytometry

A broad range of analytical techniques are implemented in pharma R&D, such as chromatography, spectroscopy, and flow cytometry, to aid the development of successful pharmaceutical compounds. Flow cytometry has clear advantages for drug discovery research, particularly by facilitating rapid drug molecule screening, but its high throughput, multiparameter functionality is hampered by the immense output of highly complex data. significant expertise is required to interpret this data correctly, and there is a lack of standardisation in assay and instrument set-up.

Flow cytometry case study 

The challenge 

The senior director of manufacturing for a mid-sized biotech company specialising in autologous cell and gene therapies was becoming increasingly frustrated with the time, effort and resources that his team of senior scientists were spending on manually analysing cytometry data to assess the quality of transduced cells. Added to this, strict validation protocols meant that if manual gating were not within a variance threshold, then it would have to be repeated. The result was more time spent on manual analysis and time wasted on repeat analyses. 

Traditional software solutions were not fit for purpose as multi-step gating had to be carried out manually, which was time-consuming and highly variable. They attempted to build in-house tools but could not develop them to meet the high regulatory standards required.

Solution 

The Aigenpulse Platform offers a next-generation approach to software in the compliant space. The Aigenpulse Platform [GxP] with the CytoML suite can be rapidly deployed on-premises or in the Cloud and enables streamlined automated cytometry analysis at scale and leveraged ML-assisted data processing. Using the Platform, the team was able to quickly configure their regulatory-required gating and analysis strategies – including multi-step gating, isolating sparse populations and representing data in lower dimensions using tSNE. 

Impact 

Now, this team can automate analysis and minimise manual input on all of their transduced cell products using their configured pipelines. Repeated analysis fell from as much as 50% to zero. The overall throughput enabled the equivalent of every QA scientist to process up to five times more data per day. As QA reports are automatically generated for every analysis, this ensures that the organisation complies with recording and logging every data processing and analysis step for every patient cell batch.

Looking ahead

Information integration is playing a major role in breaking through the barriers to lab innovation and, as a result, there is a significant transformation underway in the informatics tools available to integrate the solutions so that data is no longer inaccessible in a single purpose system. Until recently, pharma labs scaling up their automation capabilities faced a roadblock when it came to data processing and analysis. Now, researchers have the tools to generate information with higher transparency, reproducibility, and quality to expedite the discovery and development of tomorrow's drugs.

^References

^{Total Lab Automation Market - Growth, Trends, and Forecasts (2020 - 2025), Mordor Intelligence, 2019.}

^{Breaking Through the Barriers to Lab Innovation, Technology Networks, 2015, https://www.technologynetworks.com/informatics/articles/breaking-through-the-barriers-to-lab-innovation-184257}

^{Implementing FAIR Data Principles: The Role of Libraries, Liber Europe, https://libereurope.eu/wp-content/uploads/2017/12/LIBER-FAIR-Data.pdf}