Surfing the Waves in Upstream Bioprocess Development – Molecule to Market Podcast with Culture’s CEO, Will Patrick
“I can tell you; this will be blowing the minds of people. I have never come across a model like this in the sector.” – Raman Sehgal, podcast host
It took SARS-Cov2 a few months to revolutionize office work. Virtual communication and cloud-based platforms are the new normal and have increased both flexibility and productivity. The cars we drive and the planes we now board again were built with heavy use of computer-aided design (CAD) and mechanical design automation (MDA), technologies, which both rely on electronic design automation (EDA). GPT-4 just scored 1,410 points in the SAT, which qualifies for acceptance at ivy league colleges.
“I can switch on the air conditioning in my car from my phone. As a consumer, we are so used to navigating technology and controlling our devices remotely just from the knowledge we get day in day out. What I can see is that you have taken some of that tech savviness and brought it into the bioprocess space. This very much ticks the box of a disruptive technology.” – Raman Sehgal
Bringing amazing technologies and to upstream bioprocess development
“There are all these companies using all these amazing tools, crazy and sophisticated tools to build software and hardware that have made such a positive impact in the world at large and in so many industries. Culture set out to do that for upstream bioprocess development. There is so much more we can do faster and at lower costs with more adoption of digitization, cloud-based working, artificial intelligence, and simulation” – Will Patrick
“While we were talking, I have literally written down the words: intuitive, sophisticated, modern, customer-centric.” – Raman Sehgal
Surfing small, mid-sized and big waves
Needs in upstream bioprocess development come and go in waves. And different types and sizes of companies at different development stages face different types and sizes of waves. Culture’s integrated, cloud-based reactor platform has value propositions for all of them, ranging from running complete projects to temporarily complementing existing in-house capacity, with no need to buy reactors, build real estate and hire dedicated staff. Clients keep full control and visibility. They can run experiments as if they were on site, and multi-location teams can collaborate virtually, from anywhere in the world.
Small to mid-size biotech companies can use Culture’s platform for both preclinical process development and late-stage process optimization. The integrated infrastructure and software platform can be the ideal package for the internal team to go fast in developing the processes, either with support from CDMOs or with those vendors focusing on manufacturing.
Larger biopharma companies may have hundreds of bioreactors at different locations across the world. They may be looking for a different infrastructure to de-risk their own infrastructure. Or they like the flexibility to bring in additional capacity temporarily. The virtualization of our systems is particularly compelling when their staff is based in different locations or traveling or working remotely.
CDMOs may not have a lot of small-scale reactors available for the many different DOEs and experiments in early-state process development or late-stage process optimization or characterization. Or capacity gets flooded by too many needs from too many clients at the same time. The CDMO can focus on clinical trial and tox material downstream, while we take care of the upstream elements. There can be a nice partnership with the CDMO and its clients via our cloud-based software. Finally, large CDMOs may have R&D groups that look for a partner to develop new technologies and methods.
Becoming more predictive
“You can call it data-driven process modeling, AI, machine learning or simulation; the more effective use of data and modeling to develop, optimize and characterize processes will bring huge value to the sector. That’s super clear to everyone. It’s already starting to happen in some little pockets, and it’s going to become a more common practice that’s going to save a lot of time and costs.” – Will Patrick
“Let’s say you have a bispecific antibody and a cell-line. Modeling and simulation should get you to 80% of your process and the product and the quality you are looking for. Running experiments then becomes further training of your model rather than process development as such. The further refined model can then predict performance at large scale.” – Will Patrick
.png)