Bluebee’s coherent data strategy set to unify Illumina’s data analytics products

This month’s attention-grabbing headlines in the field of genomics, genetics, and precision medicine can undoubtedly be attributed to the acquisition of Bluebee by Illumina. I spoke with Hans Cobben, CEO of Bluebee and now Vice President of Software Platforms and Applications at Illumina, regarding the impact of this acquisition on the future of genomics data analysis and genomics applications in the clinical setting.

Bluebee, a 35 person company, headquartered in Rijswijk (The Netherlands) and Mechelen (Belgium) was founded in 2011 by a group of domain experts and bioinformaticians as a spin-off of the Delft University of Technology and Imperial College London. Bluebee has since developed a unique genomics analysis solution for research and clinical labs which is built on a robust and secure infrastructure hardened in fintech.

enlightenbio:  First off, congratulations on last week’s acquisition of Bluebee by Illumina. How do you envision the integration of the Bluebee platform – with its various capabilities – will translate into the Illumina portfolio of products with the currently existing BaseSpace Sequence Hub and DRAGEN Bio-IT Platform?

Hans Cobben: Thank you and yes, very exciting news! Conglomeration is at its fullest in the genomics sector. Bluebee of course, had interactions with Illumina before, for partnership and joint deals, but this acquisition has clearly been accelerated because of COVID-19. Bluebee offers a coherent data strategy that aligns well with the various Illumina data analytics products. Bluebee is a unifying force for all of that.

Yes, there are different sets of existing solutions in the Illumina portfolio. There is the DRAGEN Bio-IT platform, which is mainly focused on the instruments sector and the processing of raw data with a number of bioinformatics pipelines for very specific applications. There is BaseSpace broadly used by many to manage their runs on the instruments, to capture raw data, and for downstream data processing and analysis. In addition, there are a number of different applications such as Clarity LIMS, Sequence Hub, Variant Interpreter, Cohort Analyzer, or the Correlation Engine. All these products have been previously developed, or acquired by Illumina. On top of it, Illumina for quite some time has already been building an underlying architecture, which is where Bluebee comes in. Bluebee will be completing the entire offering with its technology by functioning as an integrating force for everything that is already in place. The goal is to provide uninterrupted service to existing customers while this integration takes shape.

EB:  At Illumina you will have the new role of VP of Software Platforms and Clinical Applications. What are some of the exciting aspects of your new role? And what will your main focus be? What will happen to the Bluebee team?

HC:  I will predominantly focus on unifying the various solutions into one single architecture delivered via one single, delightful user experience. And of course, via the Illumina infrastructure we will further aggregate, share, and explore small and large amounts of genomic data for both translational research and for clinical use which is becoming more relevant as we are gradually moving applications from the research lab into the clinic. Combined we will have a team of about 200 to bring working together in bringing the Illumina data strategy to market. It’s all very exciting. I will be owning all the software platform products, except for the DRAGEN platform which is owned by Rami Mehio.

Everyone from Bluebee will be integrated into the Illumina organization – everyone in California, Belgium and the Netherlands. This includes the developers, products specialists, and sales force. We see this integration as a compliment for the product we have built. In the genomics market Illumina is one of the most important and biggest players with over 7,700 employees. One of the reasons that attracted Illumina to Bluebee was our capital efficient way of building the platform and the company.

EB:  Any thoughts on the timeline to see such an integration fully in place?

HC:  I’m rather impatient, and I think it shouldn’t take a year.

All of the software/platform products will be supported for the existing customers going forward, so no one will be disrupted. But gradually, we’ll put it all together on a single platform. The result will be business as usual for everyone, but more and more exposure of the newly integrated solutions and services for those who are currently customers, and of course for new customers.

It will be quite an aggressive path, at a pivotal point in time where we can really start offering integrated solutions for biopharma, integrated solutions for population genomics, and integrated solutions for clinical applications. There is definitely so much room we can start occupying, and that’s exactly what we will do.

Lastly, under that integrated Illumina umbrella, a vast, global sales force will be part of it.

EB:  How important was the fact that the Bluebee platform supports multiple, globally positioned clouds and not just the Amazon cloud?

HC: Indeed, very important. We are offering a multi-cloud, global, and compliant platform. With that we have an extremely scalable architecture which will give users access to streamlined data processing – we don’t shy away from large data volumes. The combination of these factors was definitely very attractive to Illumina.

EB:  How will this acquisition impact the entire genomics sector and other companies in the sector?

HC:  It’s time to make an industry out of the domain. That being said, the question is who is in the position to do that? You cannot do that if you are a startup even if you’re a startup with let’s say 100+ people. For the industrialization component to take place you need to be a bigger and well penetrated player in the market. Bluebee alone can’t do that, but now we are Illumina, and this is going to be essential for the industrialization strategy, the standardization, and the credibility of new solutions being released to the market. There is, of course, some competition in the sector and that’s okay. In fact, if there weren’t any competition, we would be in the wrong sector.

It’s not really about technology. For example, I don’t know what’s in my portable Mac. I don’t know what kind of CPU, memory, or brand specific units are in there, and how they communicate with each other. The same applies to industrializing the genomics industry. Instead…..

“…it’s about servicing an audience that is in dire need of actionable solutions.”

Focusing on the ultimate usability of genomics for translational research and in the clinical context is of integral importance. The biggest challenge we are facing for this domain to mature is from an end-user perspective. We must integrate the end-user various components to include buying and running a sequencer, buying reagents, managing a lab, adhering to very strict and version controlled protocols, uploading the data to a machine/cloud for data analysis, programming the various aspect of data analysis, getting the result, and reporting actionable findings.

What needs to be created is a unified service manageable from within one user-friendly single application either next to a sequencer or remote from any device. One has to be able to control, monitor, and quality assure the data flow and testing. Again, the underlying infrastructure does not need to be understood by the user as long as a good user experience is created which is seamlessly managed.

“For the industrialization component to take place you need to be bigger and a well penetrated player in the market. Bluebee alone can’t do that, but now we are Illumina, and this is going to be essential for the industrialization strategy, the standardization, and the credibility of new solutions being released to the market.”

EB:  COVID-19 has turned out to be quite the disrupter of our current healthcare system. Any thoughts on how COVID-19 will shape future product developments and impact the advancement and adoption of precision medicine? How will it impact health care in general?

HC:  Yes, and that’s because we’re still in a world where everyone is working in his/her corner, doing valuable interesting stuff, but it’s extremely hard to share results and new insights. The data strategy Illumina is all about is more than just, “Here is your sample, and here is your outcome”, especially from a physician and patient perspective. This has become very clear with the COVID-19 phenomena – and for population genomics and infectious diseases more in general – where there is no structured data approach in place. There are also no solutions on the market to address this issue. Everyone is doing a lot of sequencing across lots of institutions and digesting their results all on their own.

“Like in any industry you have unexpected black swans that will change the shape of what you’re doing forever. That’s exactly what COVID-19 stands for right now. COVID-19 is actually the alarm bell for infectious diseases, and sequencing is one of the ways to fight the fire.”

We all understand that precision medicine, to a large extent, will rely on sequencing a large number of individuals for biomarker detection, therapeutics development, pharmacogenomics, and diagnostics. All of this requires a larger scale, data-driven approach, and that’s exactly where Illumina is well positioned. Illumina provides the instruments, the reagents, and is well represented across many organizations. Therefore, Illumina can push solutions in this market to address these specific needs through a variety of applications. Both on the research and clinical side we need more data in a centralized place. Illumina just has to enable the infrastructure in support of these needs.

COVID-19 is just one infectious disease, but there are so many others. There still is tuberculosis, Ebola, or HIV – we still have so much to learn. We are in dire need to have a data play in place at a global level. In relation to COVID-19, sequencing is absolutely the best approach to do so. COVID-19 is the trigger that has woken up the healthcare sector. If you consider the devastating effect it has on healthcare, and take into account the political and economy impact, it is enormous.

EB:  And how much of all of this is on the data side? You mentioned earlier that sharing information is a big challenge, because everybody works in their own corners, and then eventually, yes they have better outcomes, but it’s very segregated. I assume COVID has a big impact on that as well?

HC: The data side is ultimately the capstone. Without it you can’t do anything. It’s the wheels under your bike. Everything else is research, now let’s make it actionable. But you also have to interpret the data, you have to share it. Now with Bluebee’s acquisition Illumina can aggregate all this data. We can share the data between research organizations. We can accelerate everything that is research, screening, diagnostics, and ultimately biopharmaceutical development of therapeutics, or even vaccines. All of which is very relevant.

“Data analysis in itself doesn’t do anything. We need to first generate the data!”

EB: How will acquisitions and partnership in general influence the advancement of precision medicine?

HC:  What strikes me today is the genomics sector in general is not necessarily overpopulated in terms of players. Of course there was us, Bluebee. There is Seven Bridges and DNAnexus, and then you’ve got a couple of smaller niche players who are focusing on tertiary analytics, but these are also becoming more and more siloed applications. For sure, there will be further consolidation in the market. Probably Illumina will play a role, but Illumina won’t be alone. There are a couple of really big players out there that have an interest in playing a role for the genomics field to become an industry. You’ve got MGI, you’ve got Roche, you’ve got Thermo Fisher, or PerkinElmer. This big boys’ battle necessitates that Illumina elevates its position in the market. Not by going into the defensive, rather the opposite, by going into the offensive, by making the data play a natural part of its offering. It will be interesting to see how this all will play out.

EB:  What about the downstream data interpretation ecosystem? Do you think the offering of these companies are of interest to bigger players as well? Will we see lots of collaborations or acquisitions in this sector?

HC:  Most of data interpretation initiatives come out of the science corner. They’re very scientifically driven, and that’s good, but, generally speaking, I would say they have a challenge when it comes to building enterprise software. Bluebee has actually done it the other way around. First, we’ve built an enterprise software and then we added genomics knowledge. That is a totally different approach – that’s the difference between enterprise software and a vertical solution. A vertical solution (e.g. Fabric Genomics or Congenica) is a very good tertiary analysis and interpretation solution, but they often only survive by way of services, not by way of software and there is an inherent manpower challenge associated with services which are not easy to scale.

The main difference being, that with Bluebee we were able to roll out products onto the market on a global scale with 35 people, because we had good enterprise software. And in any other model, every time you sign up a customer, you need at least two specialists in support of that customer. Specialists are scarce, they are expensive, and they are a low margin. On the other hand, we used software like SAP uses software, or Oracle uses software, or anyone else uses software. We provide an integrated service which creates high value for the end user, and with a margin that allows steady growth as you go. I can’t see the growth model for very specific vertical niches, unless they get integrated in larger play model strategies.

“As a startup, you have two options: either you grow or perish. And if you cannot grow on your own, because you are in a very specific vertical niche – instead of being broad and horizontal – then you either will be bought, or you die.”

EB:  The integration of clinical genomics data with other clinical patient data has been notoriously slow. By when can we expect full integration and adoption of genomics in the clinic and where to do you think these leading developments will take place?

HC:  The consumption of clinical genomics by definition will be in the clinic, but the complexity will have to reduce substantially. That is where we were with immunotherapy five to ten years ago. Everything was very complex, very case-driven, and very much in a learning mode. But now, immunotherapy is really coming of age, with the rate of adoption constantly increasing.

On the genomics side the utilization pace is going up which is good. Eventually we’ll have a model where you get into some sort of a flying wheel effect, once it’s turning, it will turn faster, and it will go faster. So, how long will it take that is hard to say. The technology is complex, the biology is complex, the chemistry is complex, the instruments are complex, and the data processing is complex. Hence, to see the flying wheel in effect it is important to simplify the entire workflow from samples to actionable insights. I even believe it has to be almost a “push of the button” type approach.

“Again, we need to industrialize, because only if we industrialize can we get scale, and if we get scale we get commoditization. And if we get commoditization we get lower prices.”

And so what will happen is what happened with sequencing. Illumina was driving the price down, making it faster, cheaper, and hence more accessible. Now the accessibility comes down to having that data interpreted the right way, compared to the vast, public data sets out there, which are growing exponentially.

Not even one percent of the variants in the human genome are understood and known to be of any significance. We are only at the beginning! The data play is one of the driving forces in getting those numbers up. Once available and integrated in such a way that supports data exchange between research institutions, translational research organizations, clinics, and CROs, will we accomplish a vast difference in clinical outcomes.