April 2024 Newsletter-Lab of the Future

Key takeaways from this newsletter:

• Most jobs will be impacted by generative AI in the coming years.
• Digital transformation has a cultural component.
• Digital transformation will help scientists overcome lab work inefficiencies.
• Scientists will need to re-skill to support the lab of the future.
• An example of the pillars of digital transformation is presented.
• The expert knowledge of scientific theory and fundamentals will become more valuable in the lab of the future.
• Large language models (like Chat GPT for example) are  being incorporated into lab informatics software packages, which allows for natural language interaction with the software.
• Hardware for data capture will make the scientist’s life easier.

According to Cognizant, an IT services company, in the coming years 90 percent of all jobs will be disrupted by generative AI.(1)  That’s one of the reasons why I attended the Lab of the Future Conference on March 11 and 12 held at the Park Plaza Hotel in Boston.  The conference was moved to this larger venue as it sold out last year, which attests to the growing popularity of this meeting.  The focus is mainly on biopharma R&D, and that discipline seems to be at the forefront of Lab of the Future developments.  It was a well-attended event  and I would estimate that participants numbered in the hundreds. The presentations covered the subject areas of Digital Transformation, Automation and Robotics, AI, Data Strategy, Interoperability and Connectivity, and Scientific Innovation.  Thirty-two vendors populated the exhibition hall.

As stated above this is a biopharma focused meeting, but still useful for those in other disciplines. As is the case with other technology shifts, there is a cultural component that goes along with digital transformation.  You cannot simply thrust new software or technology onto a group of scientists. Mark Browosky’s (Novartis) presentation was titled “The Digital Transformation Journey-Are We There Yet?”(2)  Organizations and their staff need to shift culturally to a “data first” mentality; they must embrace data as an asset.  The scientists must accept that the software may drive the lab work in some instances, and they must trust the computational results.  Each organization undergoing transformation needs to help the scientist understand that this change in culture will help them overcome challenges in their daily work lives.

Some of the challenges to be overcome were addressed by Lee Tessler of AWS in his talk titled “Cloud as Catalyst:  Laying the Data Foundations for Future Ready Labs.”(3)  During his presentation he pointed out the current challenges faced by laboratories.  First of all, there are many repetitive, low value, manual tasks that a researcher needs to perform every day.  These tasks need to be eliminated or replaced by automation or computational methods so that a researcher can focus on what really brings value to a company:  significant value-added experimentation such as innovation for example.  Secondly, data is siloed and disconnected.  Again, going back to the presentation by Mark Browosky above, data is a company asset and should be available to all researchers in a laboratory.   Data should also be in a form that allows it to be readily reused.  Thirdly, deficient workflows lead to inefficiencies.  Disconnected lab instruments, software, and analytics can lead to collaboration roadblocks and difficulties in troubleshooting.   And the fourth challenge many labs face is that collaboration is hindered by inability to share data.

Jen Bouchard of Accenture gave a presentation titled “Digital Lab Transformation and the Workforce of the Future” in which she also addressed current laboratory challenges and solutions for moving forward.(4)   40-60% of a scientist’s time is spent on non-value added activities or inefficiencies due to paper based processes, siloed systems, lack of data visibility, and poor access to data.  These challenges echo those pointed out by Lee Tessler.  Only 38% of life science CEOs say their companies offer paths to gain skills related to AI.  Scientists will need to re-skill to support a lab of the future in which most processes are efficient, automated, and generate AI ready data.  Efforts should be made to listen to the scientists to know what their pain points are.  Tools such as AR (augmented reality) for training and voice assistants for data capture will help scientists work more efficiently.

Christopher Langmead continued the transformation theme in his presentation titled “Digital Transformation and Drug Discovery.”(5)  He discussed why digital transformation is so important at Amgen and presented their five pillars of digital transformation.  These pillars can be useful to any organization evolving digitally:

• Data generation and capture.
• FAIR data and AI/Machine Learning (ML) strategy.  FAIR data is Findable, Accessible, Interoperable and Reusable.
• Flexible computing infrastructure.  The systems need to be scalable and able to incorporate emerging and more powerful computing methods.
• Computational sciences.  An organization needs to have the right resources for AI/ML, analytics, informatics, etc.
• Data savy R&D workforce.

In addition to affecting the laboratory environment, digital transformation is also affecting the software companies themselves.

Erin Davis of Schrodinger explained how her company is using machine learning in her presentation “Empowering the Future of Collaborative, Digital Drug Discovery-From Small to Large Molecules.”(6)  Schrodinger was founded in 1990 and has developed a physics based computational platform that can be used for drug discovery and materials science.   The methods they use allow them to extrapolate into new chemistry spaces but are slow. On the other hand, machine learning is fast but only knows what you have taught it; ML requires a training set.  Schrodinger is combining the two by creating physics based training sets for ML based methods.  This approach is leading to faster generation of potential molecules for targeted drug applications.  This is interesting as very little actual lab experimentation is done until the candidates are generated via calculations.  This underscores how domain knowledge, or knowledge of science (in the physics based methods) can be used to minimize actual experimentation in a lab.  The Citrine Platform is another example of how domain or expert knowledge can be incorporated into AI modeling.(7) The understanding of physical science fundamentals and theory will be extremely valuable in the lab of the future.

Large language models (LLMs, like Chat GPT for example) are making their way into software packages. Synthace is a software company that has developed a platform for designing, running and analyzing experiments.(8)  Synthace has connected OpenAI’s Chat GPT with their platform to set up and run experiments. On the exhibition floor, Chris Grant demonstrated how a DOE for biological experiments can be generated. A scientist can explain, with natural language, the DOE they would like the software to set up, and then the AI and scientist work together to further refine the experiments.  In this way, the DOE is constructed by the software, dramatically reducing the amount of time the scientist would otherwise need to assemble a DOE. 

Sapio offers a lab informatics package to address ELN (electronic lab notebook) and LIMS (laboratory information management system) needs.  Rob Brown of Sapio gave a presentation titled “The Lab AI Revolution:  Unleashing the Power of AI and Data to Accelerate Research and Discovery.”(9)  He discussed ELaiN, Sapio’s Electronic Lab AI Notebook.  ELaiN uses LLMs to allow a user to ask it natural language questions.  ELaiN can be used to set up experiments, review inventories, find data, search reports, etc.

 A  new platform called nuVerve(10) by Rheolution will be launched in mid April and is a cost-effective solution targeting small and medium sized teams according to CEO Anis Hadj Henni.  The platform can be used to structure data so that it is available for analysis and collaboration.   The software will facilitate the ingestion and analysis of data, along with creation and automated updating of reports, so that the researcher can focus on innovation and not on manual execution of those tasks.

Other software companies of note on the exhibition floor were:

Uncountable(11)-A platform allowing for data centralization, visualization and predictive analytics.

Revvity Signals(12)-Offers the Signals Research Suite which includes a notebook for the storage and mining of data, and tools for visualizations and analytics.

SciNote(13)-An ELN which also assists in data management, project management and inventory management.

Although much of what I gravitated towards was software based, there was a good presentation on the use of hardware to make the scientist’s life easier.  Pat Leblanc of Regeneron presented “A New Era- Leveraging XR, Voice Assistance and Automation to Benefit Research Within the Biotech Industry.”(14) The focus of his talk was the TIDES (Transform Information with Digital Experimental Solutions) initiative at Regeneron.  His team interviewed researchers  in the labs to uncover pain points.  There were repetitive manual tasks being performed, and tasks were not performed in a consistent fashion.  His team saw a need to increase the quality and availability of the data.  They demonstrated technologies to the scientists with roadshows and demos in order to get their buy in.  To facilitate data capture they instituted a  technology from  Lab Voice which enables the use of voice commands to enter data or activate instruments, thus keeping your hands free.  They also used the AR (augmented reality ) of a  Hololens  (Microsoft’s AR headset) for training purposes.

So, what do you think of the Lab of the Future?  I think it’s already here.

1.  Cognizant Impact Study Predicts Generative AI Could Inject $1 Trillion Into U.S. Economy Over 10 Years - Jan 10, 2024  Cognizant Website.  Press Release. “Cognizant Impact Study Predicts Generative AI Could Inject $1 Trillion Into U.S. Economy Over 10 Years,” January 10, 2024.  (http://news.cognizant.com/2024-01-10-Cognizant-Impact-Study-Predicts-Generative-AI-Could-Inject-1-Trillion-Into-U-S-Economy-Over-10-Years)

2.  Mark Browosky, “The Digital Transformation Journey-Are We There Yet?”Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

3.  Lee Tessler, “Cloud as Catalyst:  Laying the Data Foundations for Future Ready Labs.” Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

4.  Jen Bouchard, “Digital Lab Transformation and the Workforce of the Future” Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

5.  Christopher Langmead, “Digital Transformation and Drug Discovery.” Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

6.  Erin Davis of Schrodinger explained how her company is using machine learning in her presentation “Empowering the Future of Collaborative, Digital Drug Discovery-From Small to Large Molecules.” Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

7. White-Paper-Domain-Knowledge-Integration.pdf (citrine.io)  Citrine white paper.  “Domain Knowledge Integration.” (https://citrine.io/wp-content/uploads/2021/01/White-Paper-Domain-Knowledge-Integration.pdf)

8.  Synthace: the Life Science Experiment Platform for R&D teams  Synthace website.  (www.synthace.com)

9.  Rob Brown,“The Lab AI Revolution:  Unleashing the Power of AI and Data to Accelerate Research and Discovery.”  Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.

10.  nuVerve - The Advanced Scientific Data Intelligence Platform nuVerve  (by Rheolution) website.  (www.nuverve.com)

11.  Home | Uncountable Uncountable website. (www.uncountable.com )

12.  Software Solutions for Data-Driven Science | Revvity Signals Software  Revvity Signals website.  (www.revvitysignals.com ) 

13.  Electronic Lab Notebook (ELN Software) - SciNote ELN  Scinote website. (www.scinote.net )

14.  Pat Leblanc, “A New Era- Leveraging XR, Voice Assistance and Automation to Benefit Research Within the Biotech Industry.” Lab of the Future USA Congress, Boston, Ma., March 11-12, 2024.