Finding the right balance: digital biology versus smart, transformative automation
In this latest blog by Solentim’s Dr Ian Taylor, the advent of digital biology is discussed. This is compared and contrasted with smarter automation, transformative new biology and better control over proven wet biology methods which have been developed and refined over decades.
Many agree that this is the era of synthetic biology, smarter automation and digital biology.
Published opinion shows three key areas that define the emerging discipline of digital biology – scientific data integration, multi-scale modelling and networked science1 . The conversion of biology from an analogue to a digital discipline, from manual to fully automated, brings with it many changes, both good and bad. But do we have to choose one over the other – digital over analogue, manual over automated?
Reproducibility, accuracy & data handling
Many areas of research and development could benefit from digitalisation and automation, particularly those dealing with very large sample sizes, as has already been seen in genome sequencing projects.
Fully automated labs – replacing human-driven processes with machines and using computers to monitor experiments and integrate data – can greatly enhance productivity and increase reproducibility and accuracy. Nature found recently that of 1,576 researchers surveyed, >70% have tried and failed to reproduce another scientist’s work2 – with biology facing a reproducibility crisis, changes in existing workflows and lab practices are clearly needed. Moreover, with the current drive for more assay miniaturisation and earlier decision making, this incidence of errors is only likely increase.
Moreover, as analytical instruments become more sophisticated, the quantity of data produced at each experimental stage increases, which in turn makes the need for data handling, sample tracking and informatics solutions ever more pressing.
In contrast, emerging digital biology platforms generate vast amounts of data storage and need huge data storage capabilities.
The need for iterative research, human logic & risk analysis
Despite evident benefits, many remain cautious. Biological systems are inherently complex and unpredictable, and research questions often require iterative adaptation and adjustment. This limits opportunities to standardise experimental workflows, thereby making automation challenging and validation processes time consuming. In industries where competition is fierce and time valuable, such as antibody development and gene therapy, extended validation time could cause significant issues and result in a competitive disadvantage.
It must also be considered that with new capabilities come new risks. The need for data protection and security has never been more important as digitisation and networked science become more common. There is an ever-increasing need for advanced cyber biosecurity; an emerging field specialising in predicting and diminishing risks that accompany the increased use of computers in the life sciences. There is also concern surrounding the possibility that digitalisation, and the sharing of genetic code through open access, opens the door to bioterrorism and hacking. It is evident that these new technologies and tools cannot be used without extreme caution.
Open versus Closed Platforms
Approaches based on ‘conventional’ biological research techniques have the potential to be radically improved by current and future advances in molecular biology & across chemical disciplines. This potential for improvement makes these more traditional, open platforms more adaptable and therefore, it could be argued, more future proof.
By contrast, digital biology systems can be viewed more as closed systems, in which processes and consumables are dictated by the manufacturer, potentially restricting the use of the system by the scientist. Moreover, with closed systems of this nature, there is the additional risk that proprietary elements may become scarce in supply, limiting the use of or increasing cost of the system.
The best of both worlds
With the benefits and limitations of both digital biology and smarter automation clear to see, it seems the path forward must include both the new and the old. Using digitalisation and automation to enhance current lab based processes and workflows, while maintaining human autonomy, logic and control over the process.
To employ liquid handling systems, pipetting robots, analytical instruments and data management software to increase speed and accuracy, but use human intuition and knowledge to manage the workflow, interpreting and optimising results. This approach could help enhance whole fields of research.
For example, coupling targeted DNA insertion methods and improvements in vector construction, along with developments in AI or machine learning instrumentation, could significantly accelerate the speed of cell line development in coming years.
Facilitating the change – advanced instrumentation
The amalgamation of new technology with current wet lab processes will require advanced, intuitive instrumentation. Digital biology and smarter automation are essential tools in advancing how we conduct research – to optimise the research, the tools must also be optimal.
For mammalian cell line development, these types of instruments are already in circulation. The VIPS and Cell Metric systems enable scientists in cell line development and cell engineering to establish a simple-to-use and completely integrated process, reliably isolating single cells into wells, confirming arrival and then conducting whole well imaging to assure clonality. These systems prove and document clonality, optimise clonal outgrowth and monitor cell expansion and productivity, resulting in large reductions in number of clones to screen, saving companies time and money.
In summary, the digital approach, whilst exciting and at times breath-taking to hear about, remains a high-risk strategy which currently requires significant capital investment and huge data storage capacities. Its implementation will inevitably also replace significant numbers of bench scientists. Currently, it is a strategy likely only to be risked by the biggest players. These risks could sensibly be mitigated with a strategy that enables companies to also take advantage of the transformative improvements in smarter, open architecture automation, coupled with developments in synthetic biology and improved molecular biology approaches.