Ditch Your Pipettes: Transcriptic is the Laboratory of the Future
Put down that pipette! Transcriptic is turning the costs structure of biotechnology on its head and making experimental results more reliable by using standardized, coded protocols and automated lab equipment.
Innovations in biotechnology—better pharmaceutical products, clearer understanding of the human genome, genetic modification of consumables—offer some of the most exciting opportunities to transform the human condition in the 21st century. Yet, the sector is plagued by serious hindrances to innovation, including high up-front capital costs, a crisis in reproducibility of results, and underutilization of human capital. Some firms skirt these challenges by outsourcing the actual running of their experiments to Contract Research Organizations (CROs), which themselves have become a $25 billion industry.[1] But in recent years, a new firm has leveraged digital innovation to transform and expand the value proposition of the traditional CRO.
Transcriptic, a remote, robotic laboratory, is creating new value for biotechnology firms of all sizes. Utilizing digital technology both to communicate with customers and to manage and run laboratory processes, Transcriptic fulfills clients’ experimental protocols through fully-automated laboratory services, reducing human and physical capital costs. Most importantly, Transcriptic’s use of digital technology eliminates one of the basic problems of scientific research: human errors or inconsistencies that obscure results. By using automated processes, Transcriptic minimizes human error and helps scientists get one step closer to true results.
This is how they do it. Anyone who wants to carry out an experiment—from graduate students to scientists at major firms—can access the Transcriptic website and input their experimental protocol using Autoprotcol, Transcriptic’s “formal language for specifying experimental protocols.”[2] For those not looking to write their own program, Transcriptic offers over ten pre-made protocols for industry-standard processes, such as genome isolation and editing.[3] These protocols have been tested and validated by Transcriptic’s team of scientists and engineers. For more complex experiments, customers work with the Transcriptic team to develop custom protocols. Customers then send the biological samples needed for their experiments to the Transcriptic lab, where the protocol will be completed in a fully automated laboratory.[4]
Transcriptic’s process creates value for companies by helping them run better experiments at lower costs. Experiments will be “better” because laboratory automation, particularly the use of pre-validated, ready-made protocols, limits process variability and error. Issues with reproducibility of lab results have long plagued the biological sciences.[5] By using programmed robots to complete experiments, scientists will not only know the exact protocol through which their experiment was carried out, but they will also be able to collect real-time data at every step of the process. When things do not go as planned, there will be data that can show scientists what went wrong where, allowing them to iterate on protocols more rapidly.
The model of a remote laboratory also helps firms minimize short- and long-term costs. In the short term, outsourcing minimizes up-front capital. This has the potential to transform the biotechnology industry, which has some of the highest capital expenditures as a percentage of sales of any industry sector in the United States.[6] Anyone with an idea (and the technical know-how to navigate the Autoprotocol system) can use Transcriptic’s services to carry out their experiment in a reliable, reproducible manner.
Turning these digital protocol into real experiments through the use of automated lab equipment reduces short- and long-term human capital costs. The need for skilled laboratory labor to carry out complex experiments has been one of the driving forces behind the culture of apprenticeship in the sciences that sees up-and-coming talent spending many years in the labs of star scientists, performing critical but dull lab tasks.[7] Ultimately, eliminating the need to complete “grunt work” in a lab could open up the industry to a greater inflow of talent, and allowed trained and trainee scientists to spend more of their time developing new experiments and analyzing results rather than physically sitting at a lab bench.
Despite the potential transformative power of its model, Transcriptic still faces many risks as it grows. It needs to earn the trust of its customers. Intellectual property is the life blood of the life sciences. As such, Transcriptic needs to prove that its employees and its systems can be trusted to protect information from competitors or leaks. Privacy and security requirements would be even greater if Transcriptic aims to market it services to biotech companies receiving funding from government agencies like the U.S. Defense Advanced Research Projects Agency (DARPA), which budgeted around $300 million for its biological technologies office in 2014 alone.[8] And, in an industry already worth $25 billion prior to the significant new value created by digital innovation, competition is likely to be fierce. Competitor Emerald Cloud Lab is already offering similar services. Transcriptic will need to continue to improve its processes, develop a library of validated protocols, and educate potential clients on its Autoprotocol language to maintain a competitive edge in what will become an increasingly crowded marketplace.
Word Count: 785
[1] Vassili Van Der Mersch, “Exploring the Cloud Laboratory: Advances in Biotech & Science-as-a-Service,” Nordic APIs Blog, May 17, 2016. http://nordicapis.com/exploring-the-cloud-laboratory-advances-in-biotech-science-as-a-service/
[2] Autoprotocol Website, 2016. http://autoprotocol.org/about/index.html
[3]Transcriptic Website, 2016. https://learn.transcriptic.com/crispr/
[4] For a list of the equipment available for programming in the Transcriptic lab, see https://www.transcriptic.com/workcells/
[5] See for example C. Glenn Begley and Lee M. Ellis’s 2012 study on cancer research, which found that of the 53 landmark cancer research studies, only 11% had reproducible results. (C. Glenn Begley and Lee M. Ellis, “Drug Development: Raise Standards for Preclinical Cancer Research,” Nature vol. 483, 29 March 2012, p. 531-533, http://www.nature.com/nature/journal/v483/n7391/full/483531a.html#affil-auth )
[6] Aswath Damodaran, “Capital Expenditures by Sector (US).” January 2016. http://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/capex.html
[7] Howard Gardner, Mihaly Csikszentmihalyi, and William Damon, Good Work: When Excellence and Ethics Meet, (New York: Basic Books, 2001); Jason Choi, “This Secretive Startup is Revolutionizing Scientific Research,” Huffington Post June 15, 2016, http://www.huffingtonpost.com/jason-choi/this-secretive-startup-is_b_10356182.html
[8] Sandra I. Erwin, “DARPA Biotech Office Recruiting Startups, Innovators,” National Defense Magazine Blog, November 7, 2014, http://www.nationaldefensemagazine.org/blog/lists/posts/post.aspx?ID=1662
[Cover Image] Casey Newton, “Inside the Secret Robot Lab That’s Shaking Up Science,” The Verge, December 18, 2013 http://www.theverge.com/2013/12/18/5216738/inside-transcriptic-the-secret-robot-lab-that’s-shaking-up-science
Thanks for such an interesting post! As someone who’s been through years of grunt work in a biochemistry laboratory, this is very appealing at first glace. However, it makes me think about how this will interfere in the training and development of scientists in the long-term. While the argument that scientists can avoid grunt work and focus on solving real problems is compelling, I strongly believe that going through years of repetitive trial-and-error at the bench is an integral part of the (current) path to becoming a good scientist. Do you think this disconnect between developing experiments and conducting them will hinder a scientist’s ability to think creatively about experiments? While I do see a huge benefit in industry for CROs, I do wonder whether this will have a positive or negative impact in higher education programs and research facilities that are a training ground today for scientists of tomorrow.
Thanks for the wonderful, post! I agree with ss! I spent four long years as an undergraduate research fellow in structural biology and infectious disease labs, and the core emphasis of postgraduate biological sciences training is protocol development, experimental troubleshooting, and data analysis. I fear that, if one hands off steps 1 and 2 to robots, the essence of doctoral training is close to lost. To the extent that researchers are not there to troubleshoot experiments or to understand why they may have worked (trial and error is another important part of the scientific method), they may lack the interesting questions/insights to drive their own research forward. I do, however, think that for more standardized protocols, systems like this are great. For early stage research in new frontiers in science, I’m skeptical.
As someone whose first real job was doing grunt work in a plant genetics lab, I too found this very appealing! I want to echo SS and Jacqueline’s points but I do think there are benefits for using this technology in certain labs where this work would get outsourced to technicians (or high school students like me) anyways. Plus, it’s possible that this could encourage more people to consider a career in biotech (industry or academia) who may otherwise have shied away due to the amount of time spent on repetitive tasks. Also, the money saved on labor costs could be used to fund additional research projects, which may actually have a positive net effect. I agree that this technology may have limited use in academia where scientists are expected to challenge and develop new protocols and stay close to the experiment at all times.
This technology might have more application in a commercial setting where there is a need to run the same experiment hundreds of times. For example, I could see Indigo really benefiting from Transcriptic. They could test hundreds (or thousands) more microbes for potential future development than they would be able to test by hand. They could also run the robots overnight, which would allow them to speed up testing. As we saw in the case, speed in discovering new commercializable microbes is essential to Indigo’s business model given that their business model is easy to replicate. This technology could help them move quickly from one product to the next and stay ahead of their competitors.