As technology continues to shape the way in which research is conducted, laboratory routines are adapting to the ever-changing environment
Improved integration and workflows, more structured and consistent data logging and new connectivity across the globe are setting new standards in productivity. One of the most popular digital tools and one that’s redefining the lab of the future is the electronic lab notebook (ELN). Dr Kevin Robinson spoke to Steve Yemm, CEO of BioData, to find out more.
KSR: Clearly, being able to log, store, track and, critically, reproduce data is essential in a research, development and manufacturing laboratory environment. How does the ELN help?
SY: The ELN is a vital component in the technology revolution. As we continue to progress towards Industry 4.0 — characterised by automation and data analytics — it’s vital that science keeps pace. An ELN delivers standardised data logging, faster results retrieval and new collaboration possibilities, all in a secure and compliant platform.
Organising data within a defined set of protocols rather than collecting the scattered, paper-based notes of the past is a clear benefit of using an ELN — and one that paves the way to the future of data analysis and meaningful interpretation.
Preset formats force scientists to organise their data uniformly, which not only allows for universal retrieval and analysis, but also lays the groundwork for automated searches based on structured data. It also sets a framework for the principles of machine learning to be introduced.
KSR: What are the benefits that machine learning can bring to a lab environment? Surely the point is to uncover new findings, not increase our reliance on existing data or predictive modelling?
SY: Incorporating artificial intelligence (AI) or machine learning (ML) into the laboratory is set to make a major impact in life sciences. With labs under increasing pressure to deliver against ambitious throughput targets, both in terms of volume and speed, the ability to automate repetitive tasks will deliver significant time — and therefore cost — savings.
Setting proper protocols and collating information in a consistent format creates order and uniformity in the stored data sets. As well as standardising the process at the front end of data logging, adding the principles of structured data to the back end — that is, effectively assigning metadata to each piece of data to set it in context and facilitate search — opens the door to machine learning.
The quality and structure of the data and metadata determine the speed of learning and the quality of output.
KSR: So, machine learning relies on organising data sets effectively in the first place — which is where the ELN comes in?
SY: Organising data sets is just the starting point. Once structured data is in place, machine learning can simplify experiment planning and documentation. Scientists can spend hours attempting to recreate successful experiments. Automated retrieval of the structured data already stored in the ELN can start a draft report based on past experience and will facilitate the recreation of the process … as well as a meaningful comparison of results.
KSR: Does ML only have an admin role or, as it gains traction, can the principles of ML be taken further to inform research protocols based on past experiments?
SY: Absolutely, ML offers the chemical and pharmaceutical industries huge potential. Looking ahead, experiment simulation based on past data will also be possible. Predicting likely outcomes offers significant time savings and also enables the scientist to drill down into the areas that show potential, rather than take time exploring blind alleys.
Given the wealth of data available, properly structured data allows pattern identification in the results — and the resulting models are added to the data bank. This opens up the exciting possibility of analysing and extrapolating meaning from the data to explore new experimental space that is not available in the wet lab. This so-called in silico experiment simulation offers hugely exciting potential.
KSR: I can see how universally adopting the same data format can also improve collaboration in terms of a standardised approach. Does the ELN offer any other benefits here?
SY: Collaboration between different partners, or when outsourcing agreements are in place, is facilitated by the ELN. It allows different skill sets to be introduced into the research from the specific expertise of different laboratories and scientists, all through a real-time collaboration platform wherein everyone can work on the same live data.
New features of an ELN, such as inline editing of documents regardless of file type, ensure all documentation is current and all collaborators work transparently and, importantly, from the same document version. This means that all participants can see all logged updates and annotations in real-time, with no need for emails introducing delays, different file versions and error margins. The time-saving potential of this new level of transparency is substantial.
KSR: What about security and data compliance? Are there any issues with this type of multiple access?
SY: It is possible to set different access levels, so that only a defined set of people can see and amend certain information. Many laboratories hold data that, for regulatory reasons, can’t leave the institution of source. Storing this information in a secure, centralised platform that’s accessible only by authorised personnel delivers the necessary compliance while opening up new potential for true global collaboration.
Because ensuring data integrity is vital to maintain the safety, efficacy and quality of drugs, all laboratories need to comply and, critically, be able to demonstrate their compliance. An ELN that documents activity at the time of performance creates original or true copies of data, and time and date stamping all records for audit purposes is the most efficient way to manage compliance.
KSR: Would this level of document control would also prevent data being tampered with?
SY: We’re seeing the principles of blockchain start to enter the world of the ELN, which rule out the possibility of tampering — whether inadvertently or otherwise — with archived or stored data. Although prepublication data transparency may not be possible, or is precluded for regulatory reasons, adopting blockchain in data archiving would ensure review and verification throughout the data lifecycle.
An ELN that adopts the tried and tested blockchain security measure presents solid data integrity management possibilities for the lab of the future.