Lab Manager | Run Your Lab Like a Business

Product News

Three Steps to Better Reproducibility in Research Labs

New scientist-friendly digital tools provide an integrated solution to the reproducibility crisis

by Magdalena Paluch, LabTwin
Register for free to listen to this article
Listen with Speechify
0:00
5:00

laboratory-technology-digital-assistants

Despite the innovations coming out of today’s modern laboratories, 50-70 percent of scientists report struggling to reproduce experimental results1. That’s a symptom of several different factors, from incomplete note-taking and bad handwriting to inefficient workflows, a competitive publishing environment and the high cognitive load placed on today’s researchers2,3. Put simply: insights get lost all the time amongst the fog of the busy lab environment.

We all know that reproducibility is the foundation of R&D work. At every stage of discovery and development, scientists must validate results. Validation means being able to reproduce work with precisely the same result. When results cannot be reproduced in an industry setting, this often leads to a "no-go" decision3. Lack of reproducibility can cause entire research programs to be shut down. It can also compromise funding in both industrial and academic labs4.

Get training in Making Data-driven Decisions and earn CEUs.One of over 25 IACET-accredited courses in the Academy.
Making Data-driven Decisions Course

New scientist-friendly digital tools provide an integrated solution to the reproducibility crisis. These tools can connect with each other and let scientists easily capture and access data while they work. More comprehensive data means researchers can make real-time, data-driven solutions. This drives more efficient research and better productivity in the lab. Scientists spend less time on mundane tasks and more time on the creative thinking needed to make scientific breakthroughs.

Better data capture

Scientists at the bench face a documentation paradox: it is almost impossible to accurately record results, document experimental conditions, and access data while hands and eyes are busy experimenting.

Digital tools, such as ELNs (electronic laboratory notebooks) or newer smart digital lab assistants like LabTwin, make it easy to record comprehensive notes on experimental conditions, protocol deviations, and any variations in results. Scientists can then see why and when those variations occurred. This in turn makes protocols more robust and results more reproducible.

Digital assistants are especially helpful when scientists are dealing with hazardous materials or delicate reagents, like cell lines. In these situations, researchers need to keep their gloves on while simultaneously trying to record or access data. New generation lab software can accurately record  and/or access information by voice commands so scientists do not need to interrupt their experiments to take notes or record results.

Effortless connectivity

While the first step in improving reproducibility is to help scientists capture data, the second step is to collate that data and provide information in real-time to support experimental flow.

Recent advances in artificial intelligence (AI) and machine learning mean that digital tools can now scan through large volumes of data from different sources and pick out the most appropriate information for scientists. Digital systems can accumulate theoretical expertise in experimental workflows, protocols, and scientific methods and can use this knowledge to offer critical information to scientists at any point in time.

In labs around the world, we are starting to see AI-based tools and scientists work synergistically together. AI does not have the creativity or lateral thinking that is essential to science, but it can collect and analyze large volumes of data. Humans, on the other hand, excel at planning and executing experiments if they can access the data they need.

With the aid of these new digital tools, scientists will no longer have to run between lab machines or return to their desktop and search through complex folder structures and websites to access information. Instead, with a digital assistant, they will be able to run a quick query at the bench and immediately receive results.

Real-time, data-driven decision making

Digital assistants give researchers the ability to perform any of a myriad of tasks without interrupting experiments, including sourcing lab consumables and materials, mining data from research papers and patents, identifying chemical compounds for screening projects from commercial catalogs, obtaining data from public and private databases, correlating data from various sources, or collaborating with peers. All these tasks will become part of the scientific workflow so researchers can always stay focused on the experiment at hand.

The key advantage of digital assistants is that they connect scientists to tools and information from both inside and outside of the lab so researchers can make real-time, data-driven decisions. This will lower error rates, improve productivity and ensure results are more robust and reproducible.

Conclusion

Scientists and funding agencies have been searching for answers to the reproducibility crisis5. New digital tools can help solve this problem by integrating data access and collection into scientists’ workflows. Streamlined access to information allows scientists to make real-time data-driven decisions. Digital assistants and devices automatically capture, record and analyze data, keeping records accurate so that scientists can easily maintain standardized, validated protocols and produce reproducible results. With more reliable findings, scientific advances can accelerate, bringing new products to market faster than ever before.

References:

  1. Baker, M. 1,500 scientists lift the lid on reproducibility. Nature. 2016.
  2. Eisner, D. Reproducibility of science: Fraud, impact factors and carelessness. J Mol Cell Cardiol. 2018
  3. Prinz, F. et al. Believe it or not: how much can we rely on published data on potential drug targets? Nat Rev Drug Discov. 2011
  4. Osherovich. Hedging against academic risk. Science-Business eXchange. 2011
  5. Collins, F. and Tabak, L. Policy: NIH plans to enhance reproducibility. Nature. 2014

Magdalena Paluch, co-founder and CEO of LabTwin, is a user experience designer and entrepreneur with more than 13 years experience in building new technologies. Magdalena brings her user-centric design approach, strategic thinking, and business insights to LabTwin where she leads a multi-disciplinary team of experts.